Introduction to designing crossovers without measurement

[Sven R]

Greetings,
I am making two PA speakers for a medium club live music band (my band!). They are to be 2-way speakers, with low end handled by one RCF MB12N301 – 8 Ohm, a 12¨ bass/mid-bass woofer, with Neodymium magnet. Detailed specs are attached.

High end is to be handled by a PA ribbon speaker... specs attached. Ribbon speaker is to be attached to a cast metal horn (70x30 degrees -- horizontal x vertical distribution).

I need someone who would design a complete(ly) passive crossover for this speaker (make schematics and layout) – with top notch components (no need to save money here, it would be stupid) – and perhaps also build two such crossover devices, to be installed in two speaker enclosures that I am making (each a vented enclosure with almost 40 liters of internal volume, tuned to 65 Hz, with one rectangular port).

Crossover frequency is to be 1,4 kHz. Minimum crossover slope should be 12dB/octave. It would be better if it would be 18dB/octave if no major losses are incured (12 or 18 dB per octave - debatable?).

Peak power that both those drivers can handle, short term (perhaps 100 msec?) is 1000 Watts RMS, but I shall never drive them that hard, rather, they are to have peak power at about 500 Watt RMS (AES standard, let us say), but that is tops, while they are actually going to be driven by an excellent Crown Studio stereo power amplifier that delivers two times 350 Watt at 8 Ohm. The 2-way speaker should be 8 Ohm, as seen by an amplifier. The woofer is nominal 8 Ohm, while H.F. ribbon driver is nominal 13 Ohm, but when together in an enclosure and finished, wired, they should be 8 Ohm speaker.

Crossover should have protection bulb and other nice features, but it does not have to be ¨studio Hi-Fi perfection¨. It is a speaker for club, for a rock band (not too loud band, not heavy rock band... just beautiful sounding band...).

For people who know how to do it, it should not be difficult. Learning curve is big for these free softwares and I do not plan to do any more crossover than this one. Could anyone help? Thanks for reading this, give it a consideration. Sincerely, Sven R.

[AllenB]

Originally Posted by Sven R

Crossover frequency is to be 1,4 kHz. Minimum crossover slope should be 12dB/octave. It would be better if it would be 18dB/octave if no major losses are incured (12 or 18 dB per octave - debatable?).

The slope will partly depend on how close the crossover frequency is to woofer cone breakup. The crossover frequency partly depends on how your cast metal horn pans out at its lower end.

Do you have plots of that horn?

[Sven R]

Originally Posted by AllenB

The slope will partly depend on how close the crossover frequency is to woofer cone breakup. The crossover frequency partly depends on how your cast metal horn pans out at its lower end.

Do you have plots of that horn?

Greetings,

Thanks for your response! Cone is RCF, very good one, goes up to 3k, so I think 1,4 kHz is just fine, particularly because it is a very desirable X-over frequency, since H.F. unit and L.F. unit are best ¨separated¨ at 1,4k (that frequency is also recommended by H.F. unit makers!). Here are horn graphs, as requested, in the attachment! Kindly consider my request of making this easy for me... It would be much appreciated to actually design this X-over... P. S. Also, one has to keep in mind that this is not a Hi-Fi ¨home theater¨ speaker, but, rather, a speaker for a rock band club P. A. for 200 to 400 people audience.

[AllenB]

I am happy to give you some help. There are parts of the design I would not be comfortable with unless I could measure them. For example, It looks as though your horn will lose control higher than 1.4kHz, or perhaps it will flip in this region. Otherwise there may be some uncertainty, and a no measurement approach may be as good a place to start.

[Sven R]

Originally Posted by AllenB

I am happy to give you some help. There are parts of the design I would not be comfortable with unless I could measure them. For example, It looks as though your horn will lose control higher than 1.4kHz, or perhaps it will flip in this region. Otherwise there may be some uncertainty, and a no measurement approach may be as good a place to start.

Thanks.... some of the above reasoning is definitely ABOVE my knowledge of this subject (although I am an EE, but being an Electrical Engineer and Audio person and a Musician... does not one an X-over designer make, innit, huh?)... well, I guess that is why I cannot make this X-over... so, carry on... very welcome!

[AllenB]

Maybe you could start a thread in multi-way and have the forum look at it on a different level. You could PM me a link to the thread.

[Sven R]

Originally Posted by AllenB

Maybe you could start a thread in multi-way and have the forum look at it on a different level. You could PM me a link to the thread.

So, I opened a new thread, as per this request, and all I got - with exception of ONE member (¨Chris 661¨) - was several ¨smart Alec¨ comments (waste of my time type of nonsense) and also, even by you, an open disclosure of my ¨reverse engineering¨ on a design of a well known company (which I would rather NOT have done, as I requested in PP!), and such ¨useless information¨... The only one contribution, that at least LOOKED LIKE it could lead to something useful (but it did not!), was by a member ¨Chris 661¨, as mentioned above, who, as he stated proudly, ¨spent 10 minutes¨ (of his ¨precious - $$$$$ - time¨, I presume?!) to actually use software and model one crossover with rather dubious and unusable results. Well, that is what one calls a ¨wild goose chase¨.

However, there is a silver lining, there is a wisdom I shall got back to, which I should have used from the very beginning: I shall go with BI-AMP and quit this futile excercise of trying to get a quality assistance to build a passive crossover. So, I shall ¨forget¨ about building a crossover - that is a lesson learned here. Gee, one always learns something (or, more precisely, gets reminded of an obvious solution!) even after 50 years of life as an Audio professional such as I am. Good health to you...

[AllenB]

Originally Posted by Sven R

and also, even by you, an open disclosure of my ¨reverse engineering¨ on a design of a well known company (which I would rather NOT have done, as I requested in PP!)

I did not get this from reading your PM, and I did NOT tell anyone you were cloning them, YOU did, just now.

I did try to bring interest to the thread for you, but many members here like to help people who come here to learn.

Maybe take a breather and try again, and this time don't say "I need someone who would design". I am trying to help you here.

[jimk04]

Good evening.

Thankyou for the help previously regarding my tweeter attenuation questions, good link provided there.

So here are my proposed plans.

the drivers I have are;
SB Acoustics SB15NRXC30-8 woofer
Peerless DA25BG08-06 tweeter

The components that I have come up with are;

Impedance levelling for the woofer

7.125R
2.8uF

XO woofer (1800Hz) woofer shows breakup begins up at 4k in datasheet

L=0.7mH

XO tweeter (2000Hz)

C=10uF
L=0.56mH

I used a figure of 3.75 for the calculations here much like the example in the first couple of pages of this thread as my tweeter is Znom 6ohms also. But in my previous question a few posts back from here I was asked to use the 6ohms figure in the provided calculator for tweeter damping link which gives me

Rs3.9
Rp3.3

for a 9db padding assuming I use BSC on the woofer.

Could the above tweeter confusion be clarified for me please and do the figures in general I have come up with look in the 'ball park' for a start?

Thankyou

[AllenB]

The tutorial separates the tasks of damping the impedance peak and reducing the level, which the resistors both do together.

The less attenuation you use, the less effect on managing the impedance variations. The parallel resistor was therefore chosen to be a unique value based on damping, as it has somewhat more effect on the impedance peak than the series resistor. The series resistor is the better choice as the attenuating resistor.

The main difference between the tutorial and the L-pad case, is that the series resistor is outside the 2nd order filter in the tutorial. This allows for a wider range of tweaking with this resistor and a somewhat reduced effect on the filter itself, leaving the components drawn to the right of the filter as the main concern when designing it.