Along with other resources we need an loudspeaker crossover tool box.
To correctly design a loudspeaker crossover network requires collection of data/ measurement and a means of designing and optimising the crossover functions.
This can be done in the analogue domain with passive filters or analogue active filters.
There is also emerging technology to use the digital domain as shown in this link:
http://www.acourate.com/#TUTORIALS
Prior to starting this phase of the system development it is important to have some system goals.
Once the goals have been determined and proven realistic and viable the specifications that will meet the requirements can be set as targets.
Individual components can then be selected on the basis of meeting the target specifications.
The significance of this approach is that you can specify the system around your goals rather then attemping to specify the end goals against particular components. I have seldom seen projects work out well on that basis. For diy purposes the latter might be a more pragmatic approach but it is important to have a clear understanding of what your componments will and won't do.
For example in the case of most if not all the JBL PT wave-guides they were never intended to be used below 2,000 hertz (the CM cone based wave-guides work from 200- 2000 hertz) in the JBL systems and yet we see people trying to make them go below 900 hertz.
http://www.jblpro.com/PD5000/PDF/JBL_PD5322_95_CutSheet.pdf
In this example the PT-H95HF is passively crossed over at 36 db per actave @1.5 khertz using the 2431 compression driver.
Very often JBL components are designed and specified for a particular system or application..just like any other manufacturer. If you are cloning a system and have all the components on hand then your job is mostly done.ie a JBL 4343 or 4435.
But if you are starting from almost a clean sheet of paper with perhaps one dependant or known component ie the 2435 compression driver then there is a lot more to consider. For the diy project cost and availability will impact on the above process but there is no harm is carefully assessing all the viable options.
Broadly stated my goal is:
"The idea being to bring to the table a 2 way based system that will deliver a very smooth, coherent and detailed sound not found in the earlier 43XX systems with hi sensitivity, low power compression and very low distortion. Will it be an MTM ? No. Will this be a rediculously expensive project. No.''
Breaking down that goal into specifics ; smooth response, coherent, detailed sound, hi sensitivity, low power compression, low distortion.
Without going into micro detail at this point I need to specify my sensitivity target,: What is a hi sensitivity system? 90db 1 watt, 93 db 1 watt 96 db 1 watt , 99 db 1 watt.
Why do I want a high sensitivity system?
So I can use less powerful amps of very high quality. All things being equal high sensitivity systems have lower power compression and lower distortion and higher dynamic range than the low or moderate sensitivity counterparts. This is / was clearly audible in A/ B comparisons when I visited John Nebel's wall of sound in 2004 and we compared the 4435 to the LRS6332. The LRS system was more accurate in some repects but the 4435 was more life like in its presentation imho. Its was very obvious the LRS system needed far more power for the same levels.
I think 90 db is average for a home audio loudspeaker and 99 db is very high in the PA category. A more meaningful measure is the amount of power required for a specificed acoustic output at 1 metre. A 96 db sensitivity system is only going to need 1/4 of the power of the 90 db system for the same acoutic output.
So if a 90 db system needed a 100 watt amp to produce the require dynamic range then the 96 db system would need a 25 watt amp. When you are operating full biased class A amps these are serious considerations in terms of power consumption and okay global warming!
I do have on hand a 500 watt class A, A/B amp and I propose to use this for the woofer and a smaller fully biased class A amp for the compression driver.
But seriously I tend to believe along with many other people that a properly designed systems of hi sensitivity have better retrieval of micro dynamic and spatial detail.
So I can live with somewhere in the range of 95-96 db 1 watt.
Now the other goals: smooth response, coherent, detailed sound
Fortunately a 2 way design supports the requirement for a coherent sound because there is only 1 crossover point in the most audibly sensitive part of the human hearing frequency range. Just exactly where the crossover point will be and the type of filts is yet to be determined but it will be somewhere between 800-1.5 kertz.
I think +- 3 db for the woofer and +- 2 db for the horn is a reasonable and realistic specification. Smoothness in the 1-4 khertz range is most important. I have not set frequency bandwidth yet. This will be somewhat influenced by my room, particularly room gain. However on paper I would like - 6db points of 45 hertz and 20 khertz
Below and above these points may require some augmentation from an ELF device/ aka helper woofer or a super high frequency transducer. This will be determined by the in-room response and subjective performance.
In terms of the how part and what components to specify I have drawn up a spread sheet of all the likely candidates in terms of woofers and horns.
Obviously a 90 db sensitivity woofer is not going to work within my specifications.
But if I parralleled two woofers the net sensitivity would be 95-96 db @ 4 ohms.
Lower/modest sensitivity woofers tend to have extended bass response but they are not always to well suited to reproducing the midrange region and can have a peak as a result of break-up of the somewhay heavy and stiff cones that are best suited for extended bass woofers.
This would limit crossover point options. So it will be a case by case selection if that scenario presents itself. The JBL 1200 Fe http://www.audioheritage.org/vbulletin/showthread.php?t=21713
and the Visation TIW 300 http://www.soundlabsgroup.com.au/bildgross/840_afrequenzgang.html
are options in that category.
As to single drivers of 95-96 db there are too many to discuss here but the options included a driver in prototype stage made by a local firm Loranz who claim Tad performance, the Tad 1600 series drivers if they are available, the JBL 2234 and a interesting mid bass woofer by Precision Devices PDN.15BR40 Neodymium http://www.precision-devices.com/showdetails.asp?id=79 .Audio Elegance also have some interesting 15 inch woofers using a novel motor to reduce flux modulation.
A woofer that provides easier crossover integration like the Precison Devices PDN 15BR40 and the Loranz AC-390X-B1 will have preference.
I have ruled out the JBL 1500AL on grounds of the hazards of shipping such a heavy driver and cost. ( I estimate it would cost $500 to ship each driver)
At this point the 2435 be compression driver is a given but it maybe upgraded to a large format 4 inch driver pending how the project turns out.
The type of horn or wave-guide will be critical and I propose to try at least two alternatives. I am looking a several wood horns, one of which will be CNC machined and fibreglass offerings from Azzrahorn and DDS.
When I have more specifc data and applications recommendations on some of the drivers I will post details.