Evolution of the Tapped Horn

[iand]

Originally posted by FlipC
I had come across the Transflex some time ago
while researching the original Imperial. The other was the one djk list.

Danley,
First, THANK YOU. For being the master you are and willing to pass along knowledge. I have read app 109075 a few times. Trying to get the information to soak in. My fascination with you started with the BassTech 7.

And No.
I wouldn't think they (speaker companies) would have slept on it. Why I stated that you have taken this to another level. I'm just looking at a puzzle trying to figure out the magic used to get the square piece into around hole.

Off subject:
Says your in IL but DSL is in Atlanta?

Does the DTS have pipes?
Saw a picture and it had pipes of each side of the driver. If they are are they used as ports or a type of wave guide?

And lastly
Do a McCauley 6174 sub!

I wouldn't recommend the 6174 for this type of application, the BL is too low and the cone isn't stiff enough, it's really optimised for large (15ft3) low-tuned (25Hz) reflex boxes. And even in such a box the power handling doesn't seem to be that good both from the AES shootout and my own personal experience, I fried the voice coil on one using an unlcipped 1200W amp -- replaced it with a Seismic 8196 which has lasted more than 5 years now.

If you really want to use an 18" driver -- and for a TH smaller drivers (15" or 12") are usually better, even if you need 2 or 4 of them -- there are other drivers around which are more suitable than the 6174.

Ian

[FlipC]

Wasn't that I wanted to use an 18" driver.
Nor that I wanted it to be a TH in particular. Just wanted to see what Danley could do with that particular sub. And I had assumed if he
happened to sim it up he would use it in a HT cabinet.

[Don Snyder]

Before the development of AKABAK, simulating a tapped horn was not possible.
Professionals will use AKABAK's powerful modeling tools to simulate complex
designs, while anyone can use David McBean's HornResp to quickly simulate
simple tapped horns.

~Don

[Brian Steele]

Originally Posted by Tom Danley

You need to have a “horn taper” to have the loading you need to make the response flat which the transflex didn’t have

Not necessarily, if you use a higher-Q driver and accept a non-flat upper response (the intent being to treat this with the x-over or EQ or even stuffing). See my P.O.C folded tapped-pipe here - http://www.diysubwoofers.org/project.../results2.html. It's basically a Jensen Transflex with some more folding. The upper peak disappeared when the first half of the pipe was stuffed.

Originally Posted by Tom Danley

,you need to have the driver parameters and acoustic dimensions appropriately sized which the transflex didn’t have.

True - it's response is VERY dependent on the parameters of the driver - as the "horn expansion" is no longer a parameter that can be varied.

Originally Posted by Tom Danley

It should be obvious that if these had existed before and performed like the Tapped horn does now, they would have been in widespread use all along.

OTOH, bear in mind that they did not have access to high-Pe and high-Xmax drivers in 1959 like they do now. . My little P.O.S, erm P.O.C. is 93dB/1W/1M down to below 50 Hz, but HornResp predicts it will run out of excursion in the passband with only 4W of input (I assumed 3mm Xmax).

[GM]

There's also the issue of time delay which today is a trivial matter done in the digital domain, but there still wouldn't be any marketable LF BW THs or TPs if there was only a passive filter solution.

GM

[bentoronto]

Can anybody compare the theories in this thread to conventional loudspeaker analysis?

I've been spending the weekend trying to gain a coherent grasp of the theory behind THs... with no success so far. Help!

[GM]

Tapped pipes and simple tapped horns are just TLs folded back on itself which are 6th order BPs stretched to their logical extreme and if you add filter chambers they become 8th order, same as a Karlson K15. What I perceive as a true tapped horn adheres to Prof. Leach's compression horn theory, so to understand them combined with the other is to understand TH theory, at least it seems to for me.

GM

[Don Hills]

As GM points out, "Tapped Horns" are not really horns. They are "quarter and half wave" resonators. They work more like organ pipes than horns. Some people consider them to be "transmission lines", but TLs have a very different set of design objectives to tapped horns.

I gained an understanding of the theory by using pencil and paper.
I drew a pipe with an open end and plotted the maxima and minima of the waves in the pipe for both quarter and half wave resonances.
It should be possible to illustrate the same thing with the ripple tank simulator (http://www.falstad.com/). I'll see if I can work up a model.

[bentoronto]

Gettin' warmer.

GM and Don -

1. Strange, why are there often 3 impedance peaks measured? What are the driver impedance (cone motion) peaks towards the bottom, how do these relate to wave/length/phase, and how to these relate to sound output?

2. How do you account for the high sound output (comparable the the enclosure-reinforcing levels you talk about) but found far north of them from say, 120 to 200 Hz?

Many thanks for your help... I am sure others appreciate it too.

[Don Hills]

The answer to all of your questions is... resonances. Multiple resonances.

Tom Danley's simple explanation goes like this:
For a classic front loaded "horn" (really, a pipe), you get good output at the lowest resonance, then a dip, then another peak. Adding the output from the rear of the driver to the other end of the horn/pipe excites a pipe resonance which fills in the dip.

You can see this for yourself with Hornresp. You will need:
- Hornresp.
- Volvotreter's tapped horn page: Tapped Horns

Start Hornresp.

On Volvotreter's page, scroll down to the Tangband 38Hz Tapped Horn.

Enter the parameters from his page into Hornresp and press "calculate".

If you've done it right, the SPL response picture should look exactly like his example.

Check the displacement and impedance curves too. Note that there are 3 peaks. This is because the pipe has multiple resonances, not just the fundamental.

Now remove the "tap" by changing from "Tapped Horn" to "Offset Driver".
To do this, use the input box at the lower right of the input panel. Either double-click on the "TH" until it cycles through to "OD", or double-click on the input box itself and select "Offset Driver" from the selection box.

Press "Calculate" and note that you have:
- A peak at 45 Hz
- A dip at 70 Hz
- A peak at 120 Hz.

Select Tools --> Compare Previous (or press F4), which will plot the Tapped Horn response behind the Offset Driver response. You can see how the dip is filled in. The dip at 170 Hz is worse with the tapped horn, because the 2 resonances excited by the 2 drivers are out of phase at that point.

Also check the impedance and displacement curves. Note that they are quite similar for both TH and OD configurations. In other words, the TH arrangement can give you a significantly better SPL response without adding much to the stress on the driver or amplifier.

For your second question, the higher order resonances cause the high output at higher frequencies. Many people have found that when you actually build the speaker, the resonances are much less pronounced than they are in the simulations. The simulations usually assume a perfect driver (suitable for fullrange use) and no losses in the enclosure.