Drivers to mate with horn compressions driver.

[warnsey]

G'day folks,

Well I have had such an amazing response to the thread about choosing compression drivers (thanks to everyone for their advice and honest impressions), I thought i'd try my luck again.

For one thing I seem to be learning an awful lot. One thing seems to be that 1" CD will need to be crossed at around 1.2khz and will roll of most likely (horn dependent) at around 15khz.

So I basically need 30hz to 1.2kz covered and perhaps past 15khz with a super tweeter.

My criteria is I want to end up with speakers that are around at least 98db sensitive.

I was initially going to go 15" JBL 2226, however not sure if there are better drivers around. So i'm looking for opinions on drivers to cover the 30hz to 1.2khz and how best to achieve this. ie. am I better off going for something like a 15" driver crossed at say 300hz with a smaller 5" driver to 1.2khz or am I better going with a 12" driver and crossing it at 1.2khz to avoid another crossover?

And is a tweeter worth adding in with the horns?

Any advice on brands and recommendations would once again be greatly appreciated.

Cheers

[Inductor]

You don't say what you have. It's a vague question.

[tinitus]

The specs and graph of this Beyma makes me wonder

http://www.beyma.de/fileadmin/seiten...l/18P80NdE.pdf

Looks like its ok to 1000hz, but very hard to imagine with an 18" woofer with Fs of 30hz and SPL of 100db

[warnsey]

Originally posted by Inductor
You don't say what you have. It's a vague question.

At the moment... nothing. However, It is going to be Azuurahorn AH-550 and Beyma CP385ND .

In terms of cabinet size for the woofer my maximum would be 500 (W), 400 (D) and 1100 (H) mm.

[Russellc]

You might also consider a 4430 clone. With the 2235H your in 30's neighborhood. Compensation can bring up a compression drivers high end, and reduce excessive midrange shout at the same time.

Remember, the compression driver is more efficient than the woofer is.
This headroom is used to reshape the high frequency drivers response, at the cost of some of this efficiency. This is how the econowave from the other thread, as well as a bazzillion other compression drivers are treated compensation wise. Most compression drivers have a few things in common besides really high efficency. Most are a little too hot in mid-upper mid range, and begin to fall off around 14 Khz. this why the compensation is so important in some applications. The reshaping of the response with compensation can supress this hot mid range eliminating shout and at the same time extend the high end.

Back to the 4430, which has a specific crossover which will require THAT original JBL compensation for the 2344 horn to work right. Rather than fight the response of these drivers, as well the search for the 2344 horns, you could use BMS 4552 ND or the like, and easily top the high freq. performance of the 4430 in terms of extension, and done correctly,
possibly fidelity as well. If you have the drivers, numerous threads on LHF will show you how to clone the 4430 crossover, I know I am going to do this one for sure

Russellc

[Brett]

Originally posted by warnsey
At the moment... nothing. However, It is going to be Azuurahorn AH-550 and Beyma CP385ND .

In terms of cabinet size for the woofer my maximum would be 500 (W), 400 (D) and 1100 (H) mm.

I'm not keen on using a large driver up so high. Go three way and ease your integration issues.

I would suggest an Eminence 3015LF <300Hz and a JBL 2123/2012, 18Sound 10NDA, a B&C 10N, AE TD10M, some Beyma's etc before crossing to the flare.

[EspenE]

For two-way I can recommend JBL 2226 15" or 2204 12".
Both can be used to 1,2 kHz with good results, though with the driver/horn you propose, it looks that you don't need to cross higher then 1 kHz.

I'm not sure how much compensation the Azurahorns need with the Beyma driver. Do you have access to measuring gear / simulation software?

A three-way may give better results when correctly done, but is also way more complex and difficult to get right. JBL 2123 is a very good mid unit.

[warnsey]

Thanks for the replies. I will be getting a mate to do most of the leg work. I just get to pick the drivers and design etc.

[Inductor]

Originally posted by warnsey

At the moment... nothing. However, It is going to be Azuurahorn AH-550 and Beyma CP385ND .

If using that CD be careful not to bring ~fs=800Hz to the overlap frequencies of the xover. It will need a 3. order on the HF. Your Azuurahorn AH-550 is not helping very much, with the xover, because it's a 550Hz horn, good maybe for another. Also the frequency of the CD is measured with a TD250, smoother for frequencies >2KHz. There are 550Hz CDs. So, be careful not to blow the CD if not doing accurate measurements at xover frequency, or you will be sad after. Things to look for, freq. of the horn, fs of the CD, and carefully designed xover.

[ZilchLab]

With proper compensation and a CD horn/waveguide that supports VHF, many modern compression drivers (and some vintage ones) will get to 20 kHz and beyond, some more gracefully than others.

On the low end, you try to match woofer dispersion with the horn at the crossover frequency so that they integrate smoothly. 4430 used 15" 2235H woofer with 2344A "butt-cheek" 100° horn at 1 kHz. That frequency is a "push," even for so competent an extended-bass woofer as 2235H, and illustrates the sort of compromise that is common in building two-ways.

It also illustrates the major pitfall inherent in the approach you are taking to this project. As you've seen by now, there are hundreds of tweeters, compression drivers, horns, waveguides, and woofers to choose from. How do you pick?

Two-ways are a substantial challenge, even for the most skilled and knowledgeable designers. You cannot just select random components from specs and expect the outcome to work worth a whit, let alone optimally, in combination.

Answer: Unless you are willing and able to design at the level many members of this forum do, i.e., trying different combinations and using sims and measurements to optimize the filters which integrate them into a complete system, you're not going to get there except by building a known, successful design.

Your statement above that you want to pick the components and hand them off to someone else to build the system defines an approach that is destined to fail. You've got to WORK with this stuff using the requisite knowledge and tools, as others here do, to make it happen....