Collaborative Tapped horn project

[AndrewT]

Originally posted by MaVo
Well if one made each pipe 1 meter long and did the expansion in 6 steps what would you get?

Originally posted by AndrewT
85Hz has a wavelength of 4m, leaving the quarter wavelength at 1m.

Originally posted by GM


Where did you get 85 Hz/4 m?

He asks what would you get for 1m long stepped pipes.
85Hz has a 4m wavelength and based on the info in the Danley quote, he asks us to believe that if the quarter wavelength is longer than any physical speaker dimension it equates to the theoretically correct shape.

[MaVo]

I am wondering, in which kind of implementation the tapped horn has the biggest advantage in comparison to a similar sized frontloaded horn. for example, i modeled a 40-300hz horn of both kinds, both use the same driver, have the same +-1db frequency response, the same volume, maxspl etc. here, using a TH has no benefit. can you show me cases where a TH is better (lower cutoff or more max spl or more linear response etc.) per volume than a normal frontloaded horn?

The attached picture shows both.

[tb46]

Hi MaVo: I think Tom Danley pointed out, that the main benefit for the tapped horn is size reduction. Bye the way, what is the speaker you are modelling?

[tb46]

Here is a quick re-model.

[MaVo]

Originally posted by tb46
Hi MaVo: I think Tom Danley pointed out, that the main benefit for the tapped horn is size reduction. Bye the way, what is the speaker you are modelling?

This speaker was just a test, nothing i would build. I am still researching what kind of subwoofer will be the best for my listening preferences.

Yes, Danley said that, but as my example showed, the size reduction is not allways there. There have to be some paramter combinations, where the TH benefit over a normal horn is maximised and i want to know which that are.

Originally posted by tb46
Here is a quick re-model.

Thats not really comparable, as i used 2 drivers and you one. But your example shows, that one is sacrificing the midbass smoothness for a smaller cabinet. Perhaps this is the direction in which the optimum performance of THs lies, as narrow bandwidth (2 octaves) low frequency transducers. The more you approach the normal frontloaded horn geometry / size, the more bandwidth you gain and the smaller is the difference between TH and normal horn.

[tb46]

Hi MaVo: I agree it is obviously a subwoofer. Sorry for overlooking the dual drivers. Here is an example with 2 drivers, and with some more work the size can probably be decreased a little more, but this is already quite "small".

[MaVo]

Thanks for the participation tb46, it seems - as you pointed out, that the interesting thing in THs is the ability to make the compromise of having a smaller bandwidth and getting the bonus, that a smaller volume can reproduce it with good response, all in comparison to a normal horn.

The only thing i want to get rid of, are the extreme high q peaks in the octave above the passband, to minimise the distortion of the horn, which will be to a big part, directly above the passband. So, if we achieve a TH with a falling response above its passband, it will be considerably lower in distortion.

As cowanaudio pointed out, the peaks may in reality be smaller than in simulation due to various "losses", but still, a falling response will be better.

I assume, the magic lies in choosing the right type of driver. This very high fs, high bl driver i used, has a really high cutoff, and is not practical for a TH sub. For the right driver, a high LE may help, as it lowers the R/L corner and so the upper cutoff. High mms, low fs, medium qes drivers would be probably the right choice for a lower frequency oriented small bandwidth horn. something as the lab12 perhaps?

[tb46]

Hi MaVo: quote Post #1137: "I assume, the magic lies in choosing the right type of driver."

I agree. William Cowan seems to be the expert on that subject (besides Tom Danley, of course). But, it is also a systems design challenge. Yesterday I found a thread on the Klipsch Forum: "Danley Labs DTS-20 On The Way." This is a great thread pertaining to a DTS-20 installation (and major home remodelling), and contains a few internal pictures that confirm what I had assumed before, namely that Tom uses resonators and passive filtering to tame some of the major problems. Even if these big peaks can be reduced with parametric filters and a high order low pass, it still makes a lot of sense to do what you can do relatively easily right at the speaker level. It might also make the integration with the main speaker system easier.

[MaVo]

Those interior DTS20 photos utterly confuse me! What purpose have the pipes?

[cowanaudio]

G'day MaVo

They are resonators tuned to the same frequency as the first few response peaks. You'd probably find that each branch is 1/4 WL of each peak in question. The ~5mH series inductor flattens the response and removes some of the higher frequency stuff (It's DCR doesn't hurt in many cases, too).

Cheers

William Cowan