Horn Loaded Bass Speakers

I want to build some large front loaded bass horns to fill the range from 20 - 200 Hz. I have a few questions for the horn loving folk out there:

1) What large, high quality (10" - 15") cone drivers would be suitable for such a speaker. I have read that drivers suitable for horn loading have Qts < 3 and Fs/Qts > 120 (I believe such figures indicate a stronger motor assembly than most other drivers). These are the only parameters I have been able to source on the net, does anyone have any comments?

2) Is a tractrix contour suitable for a front loaded bass horn or should I use an exponential horn.

3) The horn will have 2 sides parralel with the other 2 expanding continuously from the horn throat to the mouth; with the desired contour. The final mouth shape will be rectangular with approximate dimensions (at this early stage) of 1350mm x 500mm. My question is: Is it better to have the long dimension of the horn horizontal (as in many horn loaded mids/tweeters/woofers such as the Altec A7 Voice of the theatre speakers) or vertical (As swith many of the rear loaded full range horns intended for lowther drivers)? Is there any difference to the perceived sound stage at these low frequencies (ie. can the way the sound waves propagate to the listener change the way the music sounds).

Hope someone can help with these perhaps obscure questions


1) You'll want a comparatively stiff suspension, but I've not investigated particular specs, so I can't reasonably say more than that.
2) Despite what some might say, you can use any flare you like for a horn. Conical, tractrix, exponential, straight...you name it. Either tractrix or exponential will be fine. The difference between the flares doesn't show up in terms of whether the horn is used for front (or rear), but in terms of the length of the horn and the distortion it imposes on the signal.
3) Dispersion won't be your problem, low frequency extension will. A horn that size is only going to be good down to about 80-100 Hz. The low frequency cutoff for a horn is directly related to the size of the mouth of the horn. A horn good down to 20 Hz will be *very* large indeed. There's a reason people don't use horns for subs; it goes beyond SAF--they're just too big to fit into an average listening room. The biggest horns I've ever seen/heard were at an Edgar (possibly Johnny?) Winter concert years ago. Rectangular mouths about 10-12' on a side, and very long (they weren't folded). Seriously huge by most peoples' standards, but they had an entire coliseum to play into, i.e. lots more room than the average audiophile has available.


P.S.: In case you were wondering how they moved the Winter PA horns--they had flanges about every three or four feet of length. The whole thing unbolted, and the sections nested inside one another quite neatly. Clever solution to an ugly logistical problem.

All of the formulas I have seen incorporate a size factor when calculating the mouth size.


SF = 1 (Free air)
SF = 2 (Against a floor)
SF = 4 (Against a floor and a wall)
SF = 8 (In a corner)

These size factors reduce the required size of the horn by the factors indicated. Hence for wall/floor mounting (what I intend to use) the mouth of the horn can be reduced in size by a factor of 4.

I am not sure if this theory is sound, it's just what I've read. I think the reasoning is that 2 walls will act as an extension of the flair of the horn thus increasing its equivalent size. Do you have any comments on this?

Yes, the wall will act as extension of the horn, but...
This is one of the bugaboos about horns. By nature, horns are long. By nature, audiophiles are resistant to having the back end of their horns in the middle of the front yard. What to do? Fold the horn. Problem solved. Oops. Well, slight problem. Folding horns introduces distortion due to the discontinuity around the fold; the pressure wave doesn't propagate well around the corner. Part of the difficulty is maintaining a consistent flare rate while changing direction. Any time you change the flare rate, you introduce more distortion above and beyond the basic horn distortion mechanisms at the throat and mouth. Which leads us to the wall (and/or floor). Unless you're planning on a straight (conical) horn, you're going to change flare rates once it hits the wall. And if memory serves, conical horns produce the highest distortion of any flare rate. (Unless you plan on reconfiguring your wall into a tractrix curve...now, *that* would be something to gladden the heart of those prone to liking modern art...).
Note also that while a corner gives you another wall at 90 degrees to the first one, there's no flare rate that I'm aware of that gives you a 90 degree straight-sided angle. If you choose a wall, you've entirely lost control of the *other* side of your wave front, as there's no surface there at all.
There's another problem with using the room as part of the horn. The response will change with your listening position. That is, since you're effectively sitting inside the horn, the closer up you sit, the less of the room can be counted as 'horn.' Granted, if you're sitting in a fixed spot, this can be calculated, I suppose, but I recall seeing a paper once upon a time that went into depth on this. Unfortunately, I don't remember all the points they made, nor do I remember where to tell you to find the paper.

OK, Hows this?

2 (stereo) front horns each of a square profile with tractrix contour with mouth of 1.6m x 1.6m (mouth area = 2.56m^2) and length of approx. 1.25m (unfolded). I estimate if these are flush mounted with the listening room wall surface at an angle in each corner of the room (in plan)that these will get down to 30 - 45 Hz.

Any thoughts?
Assuming that I'm visualizing what you're describing properly, that sounds like a better plan. For some reason, 1.25m length sounds short to me, but I'm just shooting from the hip; it's not as though I've walked through the math to verify the measurements.
Something to keep in mind is that horns are only good for about 2 octaves before distortion starts rising. In the real world, commercial designs generally push closer to 3 octaves, just because it'd be really obnoxious to have five or six separate horns to cover the audible range. If you're shooting for 30-45 up to 200Hz, you should be running about 2 1/2 octaves, which should keep things fairly tidy in that respect.
If you going to want anything below 30Hz, it may be difficult to mate subs to these horns.


The 1.25m length is not a guess, This is how I came up with it:

I calculated horn mouth area from the standard formula.

Let the horn throat area equal the equivalent cone surface area of the driver (in this case a 12" woofer).

I fed the parameters into my own tractrix contour spreadsheet and read-off the length for the given mouth/throat areas.

Do you have any tips for good quality 10/12" woofers on the net with specs. I live in Australia and all I can get locally is car stuff or not so hifi speakers from electronics distributers.



1.25m length??

should give you a handy mid-range horn :)

one suitable 15" driver for a bass horn is 15pw40 from b&c - italy , we have very good experience with this (we have tried many other over the years....)

they have 12" drivers too , i dont have much experience with them as bass horn drivers though.

bye k madsen - cadaudio.dk
Not a horn, and probably far from what you're looking for, but here's my take on an efficient, reasonably sized enclosure with deep bass. (and don't tell anyone, just talking out of my butt here....haven't actually tried this!)

Seas makes a 10" woofer with low F's and a rated efficiency of 95db/1M. Put this in a bass reflex cabinet of your liking using your favorite box design software.

Adapt this to Wayne Parnham's Pispeaker "corner horn" design using his guidelines contained in the whitepaper on his site. It looks just like a big doghouse! As the name implies, this gets tucked right in the corner to get another 3db efficiency. (maybe you could get really crazy and stack a pair of these in each front corner for even more efficiency, drivers in parallel)

Just an idea, probably a silly one....