Smallest midrange for effective open baffle dipole application?

[Javachip]

What is the smallest midrange you would use on an open baffle to get the dipole benefits? Would two smaller drivers on an open baffle, wired in parallel, be equivalent to a single larger driver of equal radiating area?

The Linkwitz Orion uses one 22 cm (8.6") midrange. The Linkwitz Phoenix uses 2x 21cm (8") midranges. The NaOII by Music and Design uses 2x 18cm (7") midranges. Would a 16cm (approx. 6") driver be too small for open baffle dipole use? Bass will be provided by a different driver.

Thanks in advance.

[ashok]

" What is the smallest midrange you would use on an open baffle to get the dipole benefits? "

What do you mean by this ?

For a dipole the driver size and parameters determine how low it will work effectively. The driver diameter will also determine how high you can crossover. So both requirements are opposite in nature.

Larger driver will go lower but will also probably crossover at a lower frequency to the tweeter. Smaller driver will need to be crossed over higher and can also switch to a tweeter at a higher crossover frequency.

Personally I like to crossover from mid to bass driver as low as possible keeping in mind the spl capability. This essentially means that drivers like 4 inches are not for me. Some 51/4 inch drivers might be useable with a limitation on max SPL capability. Suitable 6 inch drivers would be ideal . 8 inches are great but not many will fit the bill.

Very small mid drivers tend to drive down the image due to the mid bass coming from the bass driver which is much lower down which detracts from realism. But that's what I found. You could always try out what you want and see if it suits your taste. Crossing over to the bass driver as low as you can below 200 Hz is best. Even 200Hz is not good enough for me. Around 100Hz is a good place to crossover and keep the full midrange in the mid driver. You really need a very good driver !

[454Casull]

A 6" would be fine.

[planet10]

Many are building (me included) the MJK passive OB with a 3" mid-tweeter.

dave

[Drew Eckhardt]

Originally posted by Javachip
What is the smallest midrange you would use on an open baffle to get the dipole benefits? Would two smaller drivers on an open baffle, wired in parallel, be equivalent to a single larger driver of equal radiating area?

The Linkwitz Orion uses one 22 cm (8.6") midrange. The Linkwitz Phoenix uses 2x 21cm (8") midranges. The NaOII by Music and Design uses 2x 18cm (7") midranges. Would a 16cm (approx. 6") driver be too small for open baffle dipole use? Bass will be provided by a different driver.

Thanks in advance.

It depends on your output level requirements, bass cross-over frequency+slope, and baffle size which determines where the dipole roll-off starts. The Orion is the minimum which will provide reference level home theater output and keep up with the tweeter crossed in at 1440Hz 4th order at the top end with a 4th order acoustic 120Hz high-pass filter. Two 18cm drivers as used in the NaO have about the same surface area. The Phoenix drivers were required to accomodate a 100Hz second order acoustic cross-over on the bottom end.

Lots of us rank the Orion mid-range as the best ever or within the top 2-3 speakers we've heard. One can speculate that's because it covers the human vocal range and most instrument's fundamental frequency without severe changes in directivity.

Having played with my Orions and a separate sub-woofer, I think you want to maintain dipole radiation below 80Hz.

[gainphile]

I have extensively tried 5" (P13) and 10".

5" is acoustically too small for a decent sized baffle to accommodate bass. But they have the best dispersion. Mounting a 5" on narrow baffle (let's say 19cm) gives best midrange. On a 30cm baffle (which accommodates 10" woofers) the mid response is not smooth.

10" gives decent bass and on 30cm baffle the dips are above 1k, which is good. But it beams severely at 1.5kHz and is not nice.

So that leaves 8" as possible "best of both world" in terms of dispersion and bass capability. But I have not personally tried them.

[Javachip]

Thanks to all for the helpful comments.

Drew, do you find it surprising that the Orion has such a good midrange performance, with the 22cm magnesium cone running up to 1440 Hz? To me, that seems to be pushing the envelope. I would think that uneven dispersion (beaming) and lack of detail would become noticeable, when a cone that big is vibrating that fast. But I can't argue with the results. I wonder how much of the Orion's stellar midrange is due to the active EQ and FR shaping, and how much is due to the driver itself?

[Drew Eckhardt]

Originally posted by Javachip
Thanks to all for the helpful comments.

Drew, do you find it surprising that the Orion has such a good midrange performance, with the 22cm magnesium cone running up to 1440 Hz? To me, that seems to be pushing the envelope. I would think that uneven dispersion (beaming) and lack of detail would become noticeable, when a cone that big is vibrating that fast. But I can't argue with the results. I wonder how much of the Orion's stellar midrange is due to the active EQ and FR shaping, and how much is due to the driver itself?

I'd speculate that the midrange is very good because directivity through the human vocal range and most instrumental fundamental frequencies is more uniform than you get with a conventional speaker. Other speakers with uniform polar response like the Plutos I built for a small apartment sound similar. The Orions might take the next step to exceptional through very low distortion facilitated by the motor design and pistonic magnesium cone with good suppression of the resonance. There's definitely no loss of musical detail and texture - it's a lot like an electrostatic midrange with a wider sweet spot. The dipole directivity is also much more tolerant of side-wall proximity than conventional speakers, especially when toed in past the listening position (no image shift or anomalies even in a room with a wall 2' from one speaker and 15' from the other).

At lower frequencies dispersion is limited through dipole cancellation with polar response 20 log cosine alpha; -3dB at 45 degrees, -6dB @ 60 degrees, -12dB @ 75 degrees. It widens a bit at higher frequencies due to the low-pass filter introduced by the motor + basket structure and beams a bit later than a conventional closed box would since you're picking up some of the back-wave radiation which has undergone a full 180 degree phase shift. Regardless, at 1440Hz the driver is only .7 wave lengths wide and directivity in line with what's happening at lower frequencies.

Some directivity is also a good thing here since it helps avoid the off-axis peaks above the first on-axis dipole peak.

The phase plug might help too.

You might take a look at SL's Phoenix measurements

http://www.linkwitzlab.com/sys_test.htm

The baffle width is close (12.5" on the Phoenix vs. 13" on the Orion) and driver Sd identical at 220 cm^2 on each design. The Phoenix wings may be bigger (the Orions start out at 2", narrow, but then broaden a lot as they reach the woofer baffle). The Orion has the phase plugs.

Don Maurer's W22EX001 Phoenix variant test sequence is interesting too. He characterized two dozen baffle configurations. Raw data is only given for a few. The points at which response 30 and 60 degrees off axis exceeds a +/- 2dB deviation from the mean are noted. Orion panel dimensions are included; I don't know whether that test was run with a single midrange or both drivers.

http://home.pacbell.net/donwm/PrototypeTestResults.htm

John Krevosky's dipole design page has good illustrations of the response getting wider above the first dipole peak and how driver directivity interacts with this:

http://www.musicanddesign.com/Dipole...n_baffles.html

[Javachip]

Originally posted by Drew Eckhardt


... driver Sd identical at 220 cm^2 on each design. The Phoenix wings may be bigger (the Orions start out at 2", narrow, but then broaden a lot as they reach the woofer baffle). The Orion has the phase plugs.

Is that right? Doesn't the Phoenix have almost twice the effective midrange Sd as the Orion, since Phoenix has 2 x 21cm, versus 1 x 22 cm for Orion?

[Drew Eckhardt]

Originally posted by Javachip


Is that right? Doesn't the Phoenix have almost twice the effective midrange Sd as the Orion, since Phoenix has 2 x 21cm, versus 1 x 22 cm for Orion?

The Phoenix has two 21 cm drivers each with 220 cm^2 Sd for 440 total. The Orion has one 22cm driver with 220 cm^2 Sd for 220 total.

Narrowed dispersion due to driver diameter should be no worse with the Orion's W22EX001 than the Phoenix which has more published measurements.