For faithful and accurate reproduction 3 way speakers are good enough.One can get some idea of it by listening to the Bose 201 & 301 by a/b switching.
A 'conventional' three way will operate with all three (or 4, or 5) drivers crossing over at equal spacing, so we normally see crossovers like 200-400Hz and 2-4KHz for most of them, with the midrange handling the 'mids' only.
This is a design where the wideband driver will go as much as 80Hz-10KHz or even more, and maybe assisted by a large bass driver and a small supertweeter (both are optional). Crossovers may be extremely simple, or electronic, and in some cases there may be no crossover at all in the lower region, using only the driver's natural low-end rolloff. Essentially it's a full range design, not a three way. I would hesitate to call it a 3-way, as it is quite different in its principles (and in sound).
I am actually planning a build exactly like this in the next few months. I already have the helpers, I just need to get my hands on the main actor once i have the funds.
Technically - that is a 3 way is it not?
Regards
A 'conventional' three way will operate with all three (or 4, or 5) drivers crossing over at equal spacing, so we normally see crossovers like 200-400Hz and 2-4KHz for most of them, with the midrange handling the 'mids' only.
This is a design where the wideband driver will go as much as 80Hz-10KHz or even more, and maybe assisted by a large bass driver and a small supertweeter (both are optional). Crossovers may be extremely simple, or electronic, and in some cases there may be no crossover at all in the lower region, using only the driver's natural low-end rolloff. Essentially it's a full range design, not a three way. I would hesitate to call it a 3-way, as it is quite different in its principles (and in sound).
I am actually planning a build exactly like this in the next few months. I already have the helpers, I just need to get my hands on the main actor once i have the funds.
Why not? Bose builds speakers which are fatiguing and absolutely not *accurate*, which the ever-*faithful* marketing department hypes up so much that people feel they should buy them no matter what the price.
The closest examples of this (commercial/well-known) are the Orion and the R909, where the bass speakers basically only go up to about 100Hz, or 2-2.5 octaves. This is NOT a three way design, I can tell you that. And I don't consider design like this a three way either I will be documenting the build with some detail.
As for inspiration, I have yet to research it throughly, once I see something I like, I'll be sure to post it. As of now I have not yet decided if I'm going to go it alone or derive off someone else's work. I may also totally make this a junk box project. It is a regular habit of mine, so don't hold out hope that my work is something you'll be able to work on because everything will come out of my junk box and locally available bits. And then it'll become both unique and useless for anyone else, an unfortunate combination. I'll try and make it duplicable to some extent though, and may serve as a starting point.
As it is my first priority will be the L18 and stands, the stands having monster subs built it. That is going to take quite a while, maybe four months or so, so don't expect anything before that.
What I have in the junk box are a pair of Philips AD163 tweeters and 4 12" drivers which need new cones, surrounds and voice coils. The bigger drivers will probably do bass duty, they were made locally to my specs about seven years ago and need to be rejuvenated after five years in cold storage. I have yet to crack them out of the wrapping. They are not subwoofer drivers, and not pro drivers either. They were specifically made for a low Q ported box application, so it's probably going to fit right in. We'll see. The tweeters are stock AD0163, and they work very well when crossed over very high (>4KHz).
The wideband will probably be the Alpair from Markaudio, Visaton T100 or similar, about $100 per driver. Research is still to throw up the right candidate, but I'd be looking at a low distortion unit with reasonable efficiency (I don't need >88dB), so the traditional widebands will probably get passed over (Fostex et al) for this project.