new design idea in the quest of SQ

[Nemophyle]

Hello guys

In the neverending quest to the perfect sound that everybody knows on theses forum, i learned one thing, that designing a speaker is all about compromises. A sentence you read 5 times a day.

The last times i considered somes particular areas where compromises are to be made :

- Directivity . A lots of speakers design don't even take it into account. However, some of the most renowned authorities, like linkwitz or dr.geddes agree on the fact that a controlled directivity is an important ker to sound reproduction.

- Bandwith and crossovers : Crossover introduces phase anomalies. a lot of people consider that it is a good thing not to cross into a certain range (may it be 300- 3000hz or 100hz-10000hz) so that a single transducer take care of this range where the most audio information is.

my ultimate goal is the perfect compromise between sound quality, dynamics, reproduction in a small room, dynamics, and price naturally,if i can think (painfully) of buying to JX6, building an array of jordan drivers is out of the question

I had the idea of combining the last JORDAN JXr 6HD with a 12" or 15" bass driver, the Jordan being put into a 12" or 15" waveguide.
The waveguide would not be used for the gain it produces or for crossing lower.
It would be used to provide to the design a controlled directivity. The gain it would provide in the drivers response, once equalized back to flat, would produce a significant fall of distortion in to the frequency region in wich the most voices information is (600-6000hz)
The crossover would be around 100-300 hz, in any case trying to let the JX6 reproduce almost all the music content. The compromise is to be made between power handling and crossover frequency / slopes, that may provide a better integration.

i have lots of ideas about the bass driver i could use, but this is not the important point here.

There is a large number of things i don't understand, and probably won't understand shortly about audio and the physics behind. I would just like to know if this idea makes any sense to anyone, or if it is just plain dumb ?
I may have forgotten or not taken certain parameters into consideration so please enlighten me

[simon5]

I think this is a great idea.

You could pair it with a Lambda bass driver.

[inertial]

Hi again Nemophile,
I suspect you now must be choice from SL and GL.....
The 15" woofer : dipole or box ?
Joke apart, it is an interesting project.
X-over: SDTP a la JK. Why not!
Cheers,
Inertial

[Nemophyle]

hello inertial !

yes you suspect good, but i wan't to go further, i can't use open baffle considering the size of my room, and i like fullrange drivers ,and the ability to have one driver reproducing most of the range, so i tried to make the best compromise. It is the existence in itself of a driver such as the jordan that makes it possible. As much as i love fullrangers and enjoy the musical experience with them, they're not so much about "fidelity" as i would want

they are a few approaches i could take about the crossover design, i could chose a 12" that have a very good extension in the midbass, and so make a first order accoustic ,and hear if it's really worth the shot.
But i don't like the idea of a 12" reproducing to 500hz while the little jordan do so, so there is much chance i do some FIR filtering with steep slopes around 250hz on my computer.
by the way,excuse my ignorance, it may be obvious, but what is "SDTP a la JK." ?

hello simon

i effectively heard that theses lambda drivers were mighty fine, i'm gonna take a look at 'em, and thank you, i hope there isn't a flaw in this design i didn't see, i would like some waveguide gurus to come, because it is the critical point, and the one i have the less examples to work on.

[inertial]

Ooops,
excuse me, Subtractive Delayed Transient Perfect by Dr. John Kreskowsky!
Yes, with FIR you can do "the impossible".
BTW, before the Jordans, that do you think about a cheap-test with one of the little Tang Band?
Cheers,

[Puggie]

I'm currently trying to partner Bandor 50s (very similar to the Jordan) with some Seas 8" drivers, How about looking at the Bandor 150 or 300 drivers as a match, they are built along the same principles as the Jordan (lots of similarities which are obvious when you look at the history), so would IMHO make a good pairing.

[Nemophyle]

ok inertial, i understand they are what is also called substractive filters.
anyway i'm certainly gonna go the FIR route as i have my old computer with good multioutput soundcard, that is gonna be converted into a big crossover


hey mr puggie thanks much for your advice, i did look at bandor drivers and they look very good.

however, the 150 mm is pricey, and as i wan't 10" or 12" i imagine the 300 will be like Two times more.

That and i would see that as a kind of "waste" to use the 150mm that is having perfect response to 5khz , as i don't need my bass driver to be so well controlled after 1-2khz

about your system, i read that in another topic, do you plan to go with a passive filter ? active in this case (low crossover frequency) is 10x times easier and better IMO

but i don't think the bass driver selection will be a problem at all, there are plenty of very good bass drivers around, that are able to reproduce very well the range of frequency i want them to.

If i go for fir filters i am certainly gonna use this driver :Hi vi W12 that according to the simulations may have a really crazy power handling, with bass at 45hz -3db in a sealed box, that i may convert to a bass reflex in the same volume (80liters) and then get 35hz -3db.

my "problem" if you can call it like that, is that i have no way to be sure with my limited knowledge that the empirical idea i had about putting this Jordan driver into a waveguide (i could go for a larger, 15 even 18 waveguide if it leads to significant improvements) is not plain dumb

so please if anyone feels like giving it's advice on this one

[phase_accurate]

ok inertial, i understand they are what is also called substractive filters.

They are not just the ordinary subtractive crossover. They can't be built in analog sice you need a constant frequency-independant delay. So you would have to implement them digitally as well. Their advantage over FIR is that they don't have pre-ringing and the advantages over analog subtractive crossovers are steeper slopes and no humps in the passband.

Regards

Charles

[Nemophyle]

wow this looks freakin interresting, could you please send me a mail with some details on this, as i understood you were quite knowledgeable in this area. thanks !

[phase_accurate]

John's info can be found here (he is a member of this forum as well BTW):

http://www.geocities.com/kreskovs/new-xooz1.html

Regards

Charles