Stepped tapped horn?

[Brian Steele]

I'm just wondering how much impact therer would be on frequency response, etc. if I implement a tapped horn in a series of steps, rather than a smoothly expanding cross-section?

Here's my idea - use the folding scheme I used for my "proof-of-concept" tapped pipe, but this time vary the internal panels a bit to create 4 sections that increase cross-section in steps.

For example, see image below - this is a cross-section of what the internals of the stepped tapped-horn would look like (dimensions in cm):



This is a tapped-horn I came up with for this driver: http://www.parts-express.com/pe/show...number=299-245

TH parameters:
S1=450cm^2
S4=900cm^2
L12=L34=31 cm
L23=278 cm
Total Length=340 cm

Should be good for 110dB (half-space) from 30 Hz up to 80 Hz @ 50W. The box is a bit big for a 10" though

[diy speaker guy]

One of the very first diy tapped horns was done in steps several years ago. Steps work just fine if you use enough of them, if I remember correctly, Danley said that if you use 6 or more steps the difference between stepped and constantly expanding is negligible.

[Brian Steele]

One of the very first diy tapped horns was done in steps several years ago. Steps work just fine, if I remember correctly, Danley said that if you use 6 or more steps the difference between stepped and constantly expanding is negligible.

Thanks. I can adjust the design to use six steps, but I wonder if I can get away with four? Where would the impact show up first?

A sub that can do 30 Hz @ 110dB half/space for $37.50. Alas, I doubt SWMBO will allow me to keep a sub the size of a young refrigerator in the living room....

[diy speaker guy]

There's no point in me speculating on what would happen with only 4 steps since I've never built or measured anything with steps. Instead, I defer to the following quote from Danley. This quote covers steps AND folding, so lucky you.

Picture a long wavelength, lets say it is of the highest frequency of interest, say 120 Hz. 120 Hz has a wavelength of 9.43 feet, its quarter wavelength is 28.3 inches. When the dimensions of a chamber or passageway are less than 1/4 wave in dimension, one can treat that
space/volume/compliance as a lumped element or elements.

What that means may not be clear but here is an example of it.

Imagine a horn, that was 6 meters long, with a 28 Hz flare and a 7 sq. meter mouth area. Instead of making a traditional horn, one made it out of pipes of stepped diameters, a "digital" horn. How many steps do you have to have before it works like a horn, any guesses?.

Well if one made each pipe 1 meter long and did the expansion in 6 steps what would you get? You get the exact same thing as a true exponential horn up to 150 Hz. What you see in the radiation resistance is a small wiggling around compared to the expo BUT with the right driver, that does not show up in the acoustic output. The effect of the steps does put a pretty deep notch in the response but the 1 meter step's first notch is at 200 Hz and the next at 400 Hz and so on. Using this box below 150 Hz, there is essentially no difference between the two.

What all this means is that when you are dealing with acoustically small dimensions, one has an extra degree of freedom in where one puts a given air volumes etc. (like where I "hid" the front volume) Where in a 1 inch compression driver and horn , having the rubber gasket obstructing even a tiny part of the horn throat can kill the hf response, one can get away with a significant deviation from "the book" if its acoustic dimension is small.

Sven in folding your horns, think in terms of how big is the quarter wavelength your dealing with.

What one cannot do is significantly change the total cubic volume of the interior, air is after all this curious combination of spring and mass and a "connectedness" to the outside. In making a bend, a common mistake is to assume the bend is longer than it really is. Going by cubic volume alone is close but also in a bend, the air is moving in an arc and so temporarily has slightly more mass (as the bulk of the volume is on the outer half of the radius of motion).

Also, at the frequency where the difference in path length between the inside of the bend and the outside of the bend is N 1/2 wavelengths, there are deep notches in the response. This is a result of the opposite phases recombining (and canceling out) after the bend. A good rule of thumb is make sharp bends with a short radius -or- bend angle is inversely proportional to radius). The point is again remember the "acoustic size" of what ever your dealing with. As you can see on the LAB sub the only large degree bend has a small radius (the one at the nose).

Does making the fold with a radius make a difference? Yes, it looks cool and I make them that way because something says "it should be". I have tested a couple low frequency horns with and without and to be honest it made little or no difference. Again what are the acoustic dimensions? In the LAB sub, I would suggest making at least the first two bends with the radius because it "feels right" (a non technical term I admit).

Still more complication, at the point where the wall area is a significant acoustic size, the sound pressure couples to the width mode resonance (caused by the parallel walls) which puts the first in a series of notches in the response coming out of the mouth. In this situation, there is a 1/2 wavelength standing wave with the pressure maxima at the walls and velocity max in the center. Coupling to this mode saps off energy at frequencies related to the N 1/2 wavelengths. Here, your horn mouth width (where it has parallel walls) also kind of sets your upper frequency limit
For a 21 inch wide horn like the LAB sub, the first width mode notch should be in the mid 300's which is a non issue.

[Brian Steele]

There's no point in me speculating on what would happen with only 4 steps since I've never built or measured anything with steps. Instead, I defer to the following quote from Danley. This quote covers steps AND folding, so lucky you.

Thanks for this.

I'm not interested in response above 120 Hz, so, if I'm reading Danley's notes correctly, my "stepped tapped horn" idea should work. I'm horrible at carpentry, so the easier I keep this design, the better for me.

I got the OK from my "financial controller" to purchase one or two 12" for another proof-of-concept. Hoping to step up the output a bit. So I'm looking now for 12" or 15" drivers with Fs~45-50 Hz and Qts=0.35~0.45...

[diy speaker guy]

Based on the pics you posted of your last tapped horn I'd say your carpentry skills are more than adequate to make a conical flare. In fact, I think a conical flare is a whole lot easier to make than steps. Either way, since you seem to be very concerned when the measured results don't line up exactly with the hornresp predictions I'd stay as close as possible to the hornresp schematic - no steps. Especially if you plan to use only 4 steps.

Considering the fact that this is only a POC build that is too big to stay in your living room, I probably wouldn't buy drivers specifically for this.

If you haven't already, I'd sincerely suggest that you start looking at the many tapped horn threads at diyaudio. There's hundreds of examples and ideas there, although out of the dozens of actual builds only a handful have actually been measured, and only a very few of those have been measured under "semi" ideal conditions outside.

[HareBrained]

Brian, you familiar with the W6 TH?

http://www.volvotreter.de/th.htm

[Brian Steele]

I had a look at it. The W6's sensitivity and Fs are a bit too low IMO to be the best match for the type of TH I'm considering - the W6's specs tend to result in THs with long lines and low Sd, which I think might be difficult to implement successfully in practice (lower Sd=greater risk that the driver's magnet structure will significantly alter the response).

Now, if TB was to do a version of the W6 with the same Xmax, but higher Fs (around 45) and higher efficiency, I might take more interest in it .

Also, the sims on that site appear to be for 1* PI space which, while they may be a more accurate simulation for in-home use, may actually give them a bit of an unrealistic SPL advantage compared to the results of the usual sealed/vented box sims (which I believe assume 2*PI space - I'm open to correction here). If that's the case, you need to subtract about 6dB from the peak linear response in order to make a more realistic comparision. Finally, even the double-folded W6 TH box doesn't have dimensions I'm too happy about - I prefer a box with as short dimensions as possible in every direction . This can be achieved with a stepped tapped horn folded like my P.O.C #1 with little or no adverse effect in the pass-band - see my thread over on DIYAudio for more info.

Finally, FWIW, some some sims I've done suggest that it might be a good idea to compare the TH sims against EBS alignments with the same driver - you may find that you're not really getting that much more by building a larger and more complicated box for the driver. THs may be a good way to get the best out of a given driver, but given that watts are cheap these days, it may not be the best use of the space it takes up to do so .

[diy speaker guy]

Now, if TB was to do a version of the W6 with the same Xmax, but higher Fs (around 45) and higher efficiency, I might take more interest in it .

Don't trust the manufacturer's specs. I measured my w6-1139si at 45 hz. This is pretty common with other diy measurements for this driver.

Finally, FWIW, some some sims I've done suggest that it might be a good idea to compare the TH sims against EBS alignments with the same driver - you may find that you're not really getting that much more by building a larger and more complicated box for the driver. THs may be a good way to get the best out of a given driver, but given that watts are cheap these days, it may not be the best use of the space it takes up to do so .

Mostly true, but there are advantages and disadvantages with all alignments w/ a given driver. Is any given alignment worth the effort? That's up to you.