Dipole design using the DPL 12 (1 Viewer)

[RichardHOS]

Just curious here, wanting to make sure I don't neglect any valid options. I'm looking for a speaker/enclosure design that can cover the range from ~60Hz to ~180Hz.

Simple to ask hard to answer question: what would the usable frequency range of a DPL 12 be? I'm having difficulty figuring out the high frequency extension of different drivers, and with dipole now I don't even know how to model the low end.

More difficult to ask but probably easier to answer question: what are the baffle requirements for appropriate dipole mounting? Are those highly dependant on the room, and freuqency response desired, or is it more dependant only on the driver chosen?

 

[Drew Eckhardt]

I'm building a pair of Linkwitz Orions (the cabinets will be done after I cut the side panels and paint; electronics should be done this Tuesday) and I'm starting to understand this dipole stuff.

http://www.linkwitzlab.com will answer all of your questions. Play with the spread sheets (notably spl_max.xls), read the phoenix prototypes section on the H-frame dipole woofer, and read the equalization notes. The H-frame is probably what you want.

The required dipole path length difference is a function of how low+loud you need to play with your displacement limitations, and the overall shape of the baffle is set to control polar response and diffraction effects. At bass frequencies, getting a big D in a small space is probably the primary concern.

In a dipole speaker there is a frequency (Fequal) at which its output is the same as a monopole.

Fequal = .17 * v/D, where v is the speed of sound (1130fps) and D the effective path difference for the front and back-waves arround the dipole.

Above Fequal, as the front and back waves become in-phase you approach a 6dB peak at .5 * v/D. Beyond that, there are a series of peaks and dips that make the speaker unuseable.

Below Fequal, unequalized output drops by 6dB/octave. To put it another way, compared to a monopole for each octave below Fequal displacement required doubles and power needed increases by a factor of four.

It becomes obvious that you need a big baffle to achieve loud deep bass. To fit this into your listening room, some sort of folding is required - wings, an H-frame, or W-frame. With short wings, the path length difference is equal to the wing depth plus half the width. On the H and W frames, width is set by the cabinet depth. The H-frame has the drivers aimed forward thereby allowing for higher cross-over frequencies; the W-frame requires a lower cross-over but takes less space and vibrates less.

Complications include the W/H-frames acting as transmission lines which require a notch filter to correct, and driver parameter interactions with the acoustic response you're trying to achieve (you can't work arround these with the enclosure as you do in a box speaker).

 

[RichardHOS]

ouch, seems much more involved than a sealed enclosure for midbass.

I've done some reading at linkwitzlabs about crossover design... I'll read his dipole material and be back with questions.

 

[Stephen Dodds]

I use a pair of DPL12s a side in an H Frame as part of my speakers which are somewhere between the Phoenix and Orion speakers.

The learning curve is steep for the non-technical types such as myself. For example, the path difference (D) Drew mentions above was stated to me variously as being measured from from the center of the driver at the back to the center of the driver at the front, as simply the front to back distance of the H-Frame, and as the width of the baffle if you folded the wings of the H-frame out. All of these give different figures.

Since I am using non-standard (for Linkwitz) components, I have had to ask a lot of questions and do a lot of experimenting, which is ongoing.

Fortunately, I use a digital crossover/EQ so I can try all sorts of experimenting, and I have measuring equipment.

The dipole newsgroup at http://www.topica.com/lists/dipoles/prefs is very good, and there are some dipole experts at www.htguide.com.

That said, my dipoles sounded remarkably good right from the beginning and the real difficulty is extracting that last bit from them.

To answer your specific question, 60 to 180 is an unusual range. 180 is probably too high for a W frame because of resonances, so an H frame would be your best bet. The depth wouldn't have to be that great either since you don't want to go all that low.


Steve

 

[RichardHOS]

Yeah, it is kinda an unusual range. I don't need to cover the sub range, since I'll cross that over and send it to the sub system, but the speakers I'm using as rear and side surrounds don't have enough dynamics in that midbass region IMO and can't go down to the target crossover point (60Hz to 80Hz) without a bit of help.

In case you're wondering, I'm using some smaller Maggies as surrounds. They blend perfectly with the L/C/R (also Maggies) as far as timbre. Matching a midbass to them is going to be difficult, which is why I want to find out a bit more about dipole design.

At higher levels, their useable frequency range is a bit more limited than their specs would lead you to believe. Whereas Magnepan lists the MMG -3db point as 50Hz, I've found that to be more like 120-150Hz at high SPL. If I could cover that last octave or so before the sub system takes over, I could avoid any holes in the response.

 

[Stephen Dodds]

Ah, I see. I'm not sure that the DPL12 is the ideal speaker to cover that range for the money. I was going to do a four way using some Focal 10 inch drivers which I have to cover the range 80-300. I did some tests and they were fine down to about 40Hz.

I simply used a slightly folded baffle with a depth of 4-6 inches.

However, I did use two of them a side.

I've heard positive reports about the Peerless 850146 10" used in a similar fashion.

I use a single Seas L21 (plus tweeter) in a dipole for my center and surrounds crossed at 90, so a couple of those or the new L26 might do the job as well.


Steve

 

[RichardHOS]

Thanks for the suggestions!

 

[Jack Gilvey]

I don't know if the DPL12 will work as high as 180Hz, although I'm using them as high as 120Hz (internal crossover on SACD player) without an issue I can detect.
One suggestion I'd make is to try and maintain dipole bass down to your lowest room mode (usually ~40Hz), then cross over to a monopole. This will maintain the true benefit (as I see it) of a dipole, largely negating the effects of the room.

 

[RichardHOS]

Good suggestions, but I'm not sure the planned mains would really carry enough weight to cover down to 40Hz (3.6's). I'd probably have to supplement them with dipole bass as well... the planned system is complex enough already, so I'll probably stick to letting the IB cover from 60/80Hz or so and down.

 

[Jack Gilvey]

No project is ever "complex enough".

 

[RichardHOS]

OK... so far I have planned an 8 driver stereo IB, BFD on each channel, 7.1 system using MMG's for rear and side surrounds with mid-bass reinforcement at each location actively crossed over and biamped, a pair of vertical maggies for the center actively crossed over and biamped (yep, four channels of amp for the center), mains actively crossed over and biamped. That's at least 18 channels of amplification, if I use a single stereo mega-amp for the stereo IB. I'm looking at 10 channels of active three-way crossovers.

So let's take your suggestion and go for broke! Until I actually hear the 3.6's in that room, I'll have to assume that their 40Hz output isn't going to cut it for HT duties (and would be suspect for music at high output levels). So I'd probably want a stereo dipole bass system to reinforce the mains (and center) from ~100Hz to 40Hz. That's going to require, say, four more drivers, two more amp channels, and all 3-way crossovers now become 4-way crossovers. Oh, and another BFD or two to tame the dipole bass.

Nah, I think it's bad enough already. By the time I'm half way done with the build and installation I'll probably give up and put a Bose system in there.

 

[Jack Gilvey]

I'd give the 3.6's a shot first at going down to 40 Hz, even the 1.6's are strong down there. Roll in a monopole/IB below that. Wow, what a system!

 

[RichardHOS]

You may well be correct... God I hope so! It would be a dream if the 3.6's could really handle frequencies down that low at fairly high listening levels. I'm trying not to get my hopes up though. I know from my experience with MMG's and SMGc's that the -3dB point Magnepan states is really only accurate at low listening levels. Above 1watt or so of input power, the lower octaves start trailing the rest of the spectrum by a wider and wider margin.

I'll use a nice prosound adjustable crossover for a while at first though to find what XO points are most appropriate and satisfying, and then build some hardwired and clean LR XO's with those points.

I've got some time to do research, thinking, and tinkering on the gap I'll have between surrounds (MMG's) and IB sub. Looks like dipole is at least a viable option, though it sounds like some active EQ might be needed in that case. Ah... more components for the rack!

 

[Mark Leitch]

I have had many maggies... the MMGs and SMGs are not even close to what the 3.6s can do. Not even remotely close... and you will spend a loooong time trying to get bass as nice as maggie bass (though I think the ribbon is the real treat). Prepare to be very pleased.

M.

 

[Stephen Dodds]

Heck, and I thought having active 3 way dipole mains, and active two way dipoles center and sides was complicated. I've only got 13 amplifier channels, three crossovers, and 9 channels of EQ. You win.



This isn't a DIY solution, although you could replicate it, but the Celestion System 6000 dipole sub would be ideal for your situation. If you wanted to get really fancy, you could switch your Maggies for MG10.1s (the skinny ones) and mount them on the subs.

Alternatively again, the Lambda dipole driver is meant to be good up into the midrange.

Cheers

Steve

 

[RichardHOS]

Focal, Peerless, Seas, Lambda... check.

I'm going to do a bit more studying (since I have the luxury of time waiting for the house to be built) on dipole theory and design, as well as sealed sealed theory and design. I'll also do some rigorous listening and level testing with my MMG/SMGc's to really flesh out at what frequency they would begin to benefit from supplementation.

Steve... curiosity - are you using Marchand or some other commercial XO's, or a DIY alternative? And what have you found to work well for EQ? I'd like some some good quality equilization to help smooth out the planar FR, but 9 channels could get quite expensive. Did you find a reasonably priced solution?

 

[Stephen Dodds]

I use the dbx Driverack 260 digital crossover/EQ for my mains. It can be configured as a 3 way crossover with four bands of EQ per output, plus another 9 bands (per channel)of EQ on the inputs. It cost me $679.

Then I use an ARX analog PEQ for my center and surrounds. It's a six band unit that can be configured to give 2 bands of EQ over 3 channels, which is how I use it. I paid $100 or so for it used.

I use Behringer and Rane crossovers for the center and surrounds.

Cheers

Steve