Tapped horn drivers

There is no specific or tight nit set of driver characteristics required for horn applications. If the horn is designed around the driver, nearly any driver that could work acceptably well in a reflex enclosure can probably have a horn built around it. Some people will argue otherwise, however, I've run across very few driver characteristics that were "unusable" in a horn.

In the real world, there are practical limitations. Large drivers often force horn designs into lower gain territory simply because the physical size is limited by personal needs/preference/portability/buildabilty.

I'm pretty quick with hornresp, if you want to share with us the drivers in question, I'm sure you'll get some ideas on how to go about wrapping them up in a horn from myself or others or both.

Regards,
Eric

Here's what I look for: low Vas, relatively high Fs, and Qes of around 0.3. But it's all about tradeoffs... some drivers work better than others, and sometimes you have to compromise.

What OklahomaWolf said is a good start. You've also got to look at the amount of cone area for the motor force that's available. There is also a LOT more to a proper design than a flat frequency response.

Nearly any decent driver can be made to "work" in a tapped horn of some kind. Just cause you can does not mean you should. There are precious few that work well enough for me to draw a fold.

I'll be honest - I am still learning what works and what does not. I have seen that smaller drivers can get away with a higher Qes. I think this is partly because the Mms is lower. Also - look at more than just BL, BL^2/Re is a good start. I divide this ratio by cone area, BL^2/Re/Sd. Numbers depend on units, but if in typical metric or SI units, I look for this ratio to be greater than 0.12 and less than 0.25. Even this is not a guarantee of a good driver. The cone has to be strong, and the driver has to have sufficient displacement to make the SPL you're after.

Real specs matter here. I can't emphasize this enough. Specs that come from the marketing department are often useless. If I did not measure the driver, I had better know and trust the source. Complex cabinets are not as tolerant of deviations in driver specs.

All of the output graphs in Hornresp are valuable. I look at all of them when designing a sub. Sure, I start with as flat an SPL curve as I can make happen, but that is the first thing I compromise to get the rest right. The time domain side of things matters too, and you can't exactly "see" those issues in the SPL plot.

Then, there is also the simple fact that at some point (and it will be different for every driver and horn combination) the predictions will not match the reality due to losses that Hornresp can not account for. Certainly, there are commonalities as to when things come off the tracks, but I have not developed any hard and fast rules yet. I can only afford to waste so much time and plywood building things that don't work.... Compression ratio is part of it, expansion rate is another.

Acoustic impedance and reactance are most often overlooked. They do matter. Group delay is usually not too big of a deal for subs, provided nothing is outrageous, but needs to be looked at too.

There is a lot more to it than I can distill into a single post. Like I said - I am still learning, and though I have theories, I still can not fully explain some of the things I have measured.

Some drivers are trickier to work with and will just give you less than ideally flat response. In the real world, most rooms are going to cause +/-6dB or more variations below 200hz anyway, so having a sub that had similar variations isn't actually as bad as one might think. In an ideal situation, it's sometimes possible to use a peaking response from a horn load to counteract a lull in response cause by room reflections. Worst case scenario, things aren't perfectly flat, and most of us are non the wiser.

The I8SW has some options... Trying to get it down to 20hz isn't something I would suggest as the harmonic distortion of a driver operating so far below Fs is apt to be pretty piss poor. Again, I'm not sure why the preference towards "high Fs" is being called, for. If you want to build a horn that plays down to 20hz, a good starting point is a driver that can play down there with low distortion, which is generally going to be a driver with low Fs.










That's a pretty usable response, from a pretty small box (internal volume just over ~3ft^3). High pass setup ~25-30hz to protect from over excursion, and a low pass at about 80-90hz, with dynamic reach to ~110dB in this passband makes for a pretty impressive performance from a single 8" woofer.

When I have a chance later on, I'll try to come up with some ideas for that JL you have in mind.

The lower it is, the smaller the horn can be to a point. But, you have to watch the compression ratio and make sure you're not going to tear apart the diaphragm either. Like I said, it's all about tradeoffs sometimes.

Anyway, those are just the three parameters I look at to quickly determine whether or not it's a driver worth throwing into my Hornresp database and playing around with. Whether or not I feel like folding it depends on a fair bit of other things, like how well I can make it fit my goals.

Simple answer - I try to have the biggest excursion maxima within the horn's passband somewhere near the driver's Fs, which is usually north of my desired tuning frequency.
Edited by Oklahoma Wolf - 7/23/12 at 9:46pm

Hello Pocket5s,

The parameters are on JLs site. Everything hornresp needs can be found under the "parameters" "tab."

quote:
"Free Air Resonance (Fs) 25.0 Hz
Electrical “Q” (Qes) 0.48
Mechanical “Q” (Qms) 7.1
Total Speaker “Q” (Qts) 0.45
Equivalent Compliance (Vas) 2.82 cu ft / 79.9 L
One-Way Linear Excursion (Xmax)* 0.65 in / 17 mm
Reference Efficiency (no) 0.25%
Efficiency (1 W / 1 m)** 85.9 dB SPL
Effective Piston Area (Sd) 77.8 sq in / 0.0502 sq m
DC Resistance (Re)*** 6.75 Ω"


The only thing not listed there is Le, which doesn't matter in this case.

---

Tapped horns can be approached different ways. The example I gave above with the I8SW is barely really reaching into the gain possible if the horn were to flare out much wider (much larger box volume). The design proposed, would be considered by many to be a tapped, tapered, transmission line. Work in hornresp for awhile and one starts to realise that these are all just different degrees of sophistication applied to different amounts of box volume to accommodate a driver characteristic in a desirable way. There are different ways to label many of these boxes.

The premise of "tapping" a line/horn, is actually misconstrued and misunderstood by most. Hornresp can simulate all sorts of box configurations, and in time, one must realise that the function of the "tap" is primarily for the purpose of gaining some control over the shape of the response by changing the point where wave-fronts combine constructively or destructively as they traverse the horn/line. The tap position in a line is a "tool" for one to use to fine tune the response of the horn. The same sort of ability to "shape" response via effecting the wave-front timing and summation, can be accomplished in non "tapped" designs by having the horn end (or line or port end) terminate at various distances to the listener.

The reason I bring up the above, is that, I see what appears to be a "tapped horn fever" spreading in the DIY crowd, much of it I believe is based on a false premise that somehow the "tapping" of the horn, results in some miraculous improvement over other types of box sophistication. In truth, the taping of a line or horn should be viewed as a fine tuning tool. Some drivers, are going to model better in alternative reflex, horn, line, or tapered line configurations when particular box volume constraints are in play. The tapped horn configuration, can represent the best use of available box space where certain drivers and volume limits are up against each-other. I don't think it's wise to set out with a handful of drivers, bound and determined to get each and every one into 1 type of box design.

---

With that in mind, what sort of box volume limits do you have in mind for these drivers? If you give 40-50ft^3 of space to work with, there isn't much reason not to build a 120-140+dB bass horn (depending on bandwidth/driver etc). If limited to just a few cubic feet, you'll have to live with the reality that a few cubic feet can only buy so much (if any) gain, depending on the driver using that space, and how well the space is used by the designer to maximise the potential box gain.

Regards,
Eric