thinking of building 7 way loudspeaker

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Here is attached sketch. What I want to do is speaker with one driver per octave (roughly) and no direct radiators. It will be very efficient, relatively small and inexpensive (no need for high end drivers, since each driver operates in narrow passband)
 

Attachments

  • 7way.jpg
    7way.jpg
    51.4 KB · Views: 708
I think the crossover will be a nightmare and the component count will negate any savings you may have on the drivers. One thing I have learned over 20 years of messing about with DIY speakers is that, generally speaking, the more drivers there are, the worse it sounds. For me, at least, I have had the best results from simple 2/3 way systems using good drivers.

I do like the idea though.
 
I'd reduce it to five ways and have an upperbass horn loading from 100-120 Hz with good drivers and shallow 6dB slopes (yes , walkin over each other).Without a midbass /upperbass horn loading it's a waste of time .Good 2 -ways are way more tricky to get right than horn loaded 5 ways not to mention that they always sound canned (said that listening to Audio Note E/L 2 way "resonators";)
 
I don't catch the purpose of doing that ,sorry...Is it just an exercise because you have many drivers around ? Maybe if you had an Accuphase digital Xo with infinite steep slopes...but then again you'd have to cascade some to divide the spectrum in 7 bands .Theoretically it could be done :just thinking what a FR driver can do ,the compromise between SPL/freq.linearity/phase accuracy puts a 4 way to be the best ,though difficult to obtain.
 
I don't think I'd just jump to a 7-way, I might make a 4, 5, and 6-way on the way. On the other hand, I see where you're going with these resonant cabinets...

Yeah, the crossover would not only be a nightmare, but I expect it would be inefficient and lose any efficiency you expect to gain from the driver configuration. This would go a lot better with multi-amping with line-level active or DSP crossovers. Perhaps as a compromise, have maybe three amps, each driving two or three adjacent-range drivers so each crossover at the drivers would be at most a 3-way.

As for the drawing, it bothers me to see drivers in resonant cabinets like that. Bass reflex cabinets aren't that bad, but I wonder if there would be audible effects when doing that closer to the midrange frequencies.

Thinking more along these "resonant" lines (while this approach may have insurmountable problems, it has me thinking), I wonder if there could be two resonances for each driver, much like IF transformers of radio receivers which use two slightly detuned resonant circuits to give a flat passband. Looking at the 125Hz and 250Hz cabinets, each one could have two ports of different length to "tune" it for a flatter passband over its full (2-to-1) range.

Not to discourage you - there's bound to be something to learn from this...

Going even further than seven drivers in seven frequency bands, there's this audio reproduction system made from a common device with 88 resonant frequency bands:
YouTube - Speaking Piano - Now with (somewhat decent) captions!
:)
 
Scaling a linear phase PC crossover to seven way's not hard. You'll need a VST host, multiple plugins, a certain amount of creativity in DACs and power amps, and a decent CPU to make it all go. Since it's linear phase and digital you actually have to work at it to make the crossover not sum flat. I'm not familiar with horn phase responses but resonance usually involves significant phase shift. So a stack of tuned horns seems like it might be recipie for trouble.

I started down this path with a five way dipole but ended up concluding a 150Hz/750Hz/2kHz four way was sufficient for my needs. I think there's a reasonable case for crossing in long throw subs below that for a five way and possibly jiggering around the crossovers to allow a 3/4 tweeter. But on the high end the problem is drivers themselves are acoustically large. If you want good directivity from a point source design a single driver's preferable to the multiple ones sketched above.
 
7 way

Okay, I don't get what the 125 and 250 hz cabinet designs are. They look like horn loaded designs, but I don't understand that it appears the drivers are facing a vent?

In any case, I think the interesting part of your concept is to see how cheap a person could make each component relative to good performance in the system as a whole. My projects have all been the opposite...expensive drivers and heavy cabinets.

Do you already have some of these components?
 
looks like fun if one already has a lotta stuff already laying around and potential dependent upon your bandpass section(s) success. One Karlson experimenter says the slotted K-waveguide used as vent works well. Some folks use panpipe several pvc ducts to distribute vent resonance. What parameters for the bandpass modules are you looking at? Have you done some modeling?
An externally hosted image should be here but it was not working when we last tested it.
 
The problem might be the vertical offset . Or maybe not . The problem might be the lower octaves boxes tuned (and resonating) at different frequencies .
Then offset between the drivers ...might be an issue too . If you tailor the
design to your ambient ,i.e. corner loading ,or anything else . I'm thinking of some kind of array ,suspended on the ceiling ,composed of one mold for the upper ways ,and some boxes. Since you start "moduling", like lego ...there's
no end,like putting everything but the super-tw symmetrically ,means 13 drivers ..:eek:
 
My five cents:

Keep it simple.
What looks too nice and attractive on paper, seldom materializes in the same way in hardware.

Just remember 5-stage space rocket. Never worked well. Too many things to go wrong. On the other hand, one, two and three stage rockets work very well.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.