Wiring for dodecahedron loudspeakers

[andytoddmsp]

Hi all! Hope you are well and keeping safe in these strange times!

I'm currently building a dodecahedron loudspeaker for acoustic measurement purposes, but I'm looking for some advice on wiring the 12 drivers.

I think that professional manufacturers use a combination of series and parallel wiring to prevent the resitance from getting too high / low, but I'm looking for any advice from this forum as to how many parallels and how many series connections

The speakers I'm using are 4 ohm, (12 of them, obviously ) and the amplifier I'm planning to use has an output of 2x400w @ 4 ohm.

My main concern is safety, and not setting the house on fire!

Any advice is greatly appreciated and will be repaid with beers if you are ever near Wigan UK

 

[andytoddmsp]

Should have also mentioned, the speakers are rated at 40w long-term, 60w short term.

Cheers,

Andy

 

[rayma]

Two parallel groups of six series 4R drivers is 12R.
Three parallel groups of four series 4R drivers is 5.3R.
Four parallel groups of three series 4R drivers is 3R.
Looks like the second option is best.

However, I would wire them as four series groups of three in parallel, that works better.
All the drivers will each receive the same power, due to the symmetrical connections.

 

[Mark Tillotson]

And 5.3 nominal ohms will be safe to drive from a 4 ohm capable amp.

It would be interesting to measure the impedance graph once connected up, since there are many modes of oscillation in such a setup, and the damping factor only applies to the fundamental mode - which should be the only mode that's excited, but I'm sure reality is more messy than that.

The ideal arrangement to decouple all these modes would be all in parallel, requiring some very high impedance drivers to be practical.

If you drive 6 series sets of 2-in-parallel, you can arrange each parallel set to be opposite faces of the dodecahedron, tying down the oscillation mode for each opposing pair to common-mode only.

If you drive 3 series sets of 4-in-parallel you can arrange each parallel set to be in a plane through the centre, each plane being orthogonal to the other two.

 

[ferret]

I would love to see a picture of the cabinets, if possible. It must have been a blighter to cut, much less assemble.

 

[rayma]

Something like these Design Acoustics D-12 speakers from the 70s.
https://www.rewindaudio.com/uploads...oustic-d12-speakers-rental-main-main_orig.jpg
https://worldradiohistory.com/hd2/I.../HiFi-Stereo-Review-1973-01-OCR-Page-0032.pdf

 

[stocktrader200]

11 - 4" full range should sound great with a sub for the deep stuff.
the 12th panel is where the stand goes

 

[rayma]

11 - 4" full range should sound great with a sub for the deep stuff.
the 12th panel is where the stand goes

The woofer can be on the bottom like the D-12. That stand has a skeletal structure for mounting.

 

[jjasniew]

My best buddy in college ('79) got his masters in ME working with such a speaker. He measured point source propagation and frequency response over various surfaces; snow, grass, blacktop... I recall the speaker was point source only in the 100's of hertz region, after which it "beamed". All I can remember about the drivers was that they were small Altecs and had aluminum dust caps. It was secured with two metal poles at opposite apexes and had all 12 faces populated.

I had seen / heard the DA D-12s in a shop in Schenectady NY, where I used to ride my bicycle to, in oder to hear a glimpse of what "real" stereo sounded like, when I was a kid. Sure wish I had those ears today!

Ah, memories...

 

[andytoddmsp]

Hi all, thanks for the replies!

The dodec is now built and working well! (Picture to follow as requested).

Just another quick question....

Speakers have been wired as described above (5.3 Ohm), and it's running with a 400W @ 4 ohm power amp. Am I right in thinking this will provide about 25w to each driver? (400/(5.3/4))/12?

Thanks again for all the help and input!

Andy

 

[andytoddmsp]

Picture

 

[Nico Ras]

Lovely piece of work. Pity the cable could not pass through the pole.

 

[rayma]

Am I right in thinking this will provide about 25w to each driver? (400/(5.3/4))/12?

That's right, assuming each driver is a 5.3R resistor.

 

[andytoddmsp]

Lovely piece of work. Pity the cable could not pass through the pole.

There's always MK2!

 

[rayma]

Hang the next pair from the ceiling by the speaker wires (with clear insulation).

 

[planet10]

Inspiring. Kudos.

I too have a docdecahedron project down the line, but hope to use 3D printing to construct the panels. Mine will be smaller, the drivers are the Foster version of the JBL 35mm FR used in the Apple eMac. I will hang them from the ceiling.

And i have enuff 4” Eastech 4” (sub)woofers to make 1 or 2 dodecahedrons with 8 drivers each for helper woofer(s).

Somewhere i have a drawing for the riangular panels and (had) a scheme for cutting them.

dave

 

[ferret]

Blimey

Agreed Nico. Damnably well done andytoddmsp. That's a fine job. Kudos to you. I actually thought they would be a little bigger - don't know why?
Kevin

 

[akhirmen_]

Two parallel groups of six series 4R drivers is 12R.
Three parallel groups of four series 4R drivers is 5.3R.
Four parallel groups of three series 4R drivers is 3R.
Looks like the second option is best.

However, I would wire them as four series groups of three in parallel, that works better.
All the drivers will each receive the same power, due to the symmetrical connections.

How about 8ohm? Is three parallel groups of four series is the best one? Tnks...