[Patrick Bateman]

Dipoles are all the rage in the home DIY audio world. I've mostly focused on waveguides, and I've posted dozens of projects that use them. There are a number of ways to control the directivity of a loudspeaker, and waveguides have been my thing for a few years now.

I'm sitting here listening to some dipoles I built, and it's astonishing how much low frequency detail exists in recordings, but sealed and ported boxes just steamroll right over those details. Okay, technically sealed boxes aren't the problem, the room is the problem.

On these dipoles, the notes on a bass guitar are as clear as day. Compared to a sealed box, the sealed box seems to blur the notes together; with the dipole each note is distinct. So distinct that you can hear the decay of the notes.

Drums are a revelation too; each strike on the drum is explosive and tight.

Here's a hypothesis on why the notes are so clear and distinct:

A sealed box radiates sound in every direction. For instance, the radiation from the door midbass is going to reflect off the ceiling, the floor, the firewall, and the opposite door.

A rock song has a tempo of about 120 beats per minute, or two beats per second, or one beat every five hundred milliseconds.

Because a sealed box radiates in every direction, we get reflections off the cabin, reflections that occur in the 'space' between those notes.

IE, if you had a drummer playing at two beats per second, ideally you'd want those beats to 'fade to black' instantaneously. But due to the reflections in the room or in the car, you don't get that.

Dipoles and cardioids aren't perfect, but they DO radiate about 50% less reflected energy, and this is why they're such a revelation with bass guitar and drums.

 

[2010hummerguy]

Dipole bass is my FAVORITE bass

Next dipole build will be with GR Research open baffle servo 12's in W frames. Should allow for narrower install and better response though the OB15's pictured above had no issue getting down. I miss that build but the next one will be even better

 

[CDT FAN]

Do those baffles have sides to them? It seems like you would lose a lot to cancellation with such a small baffle.

As for the GR Research. I had the pleasure of hearing a pair of Super-V's that Danny Richie set up himself with all of the correct room treatments. They sound glorious.

 

[Patrick Bateman]

Do those baffles have sides to them? It seems like you would lose a lot to cancellation with such a small baffle.

That's the idea

Fifteen years ago, a lot of dipoles looked like this. The baffles were as wide as you'd see on a conventional loudspeaker.

As time progressed, the baffles got narrower and narrower and narrower. Narrowing the baffle makes it work as a dipole to a higher frequency. IE, a 34cm wide baffle is only a dipole from about 1khz and down; a 17cm wide baffle is a dipole from 2khz and down. Making it half as wide moves the dipole transition up by one octave. (And doing it again would have the same effect.)

When you look at the measurements and the sims, they look really bizarre, because the response is anything but flat.

For instance, here's four Dayton ND91 midranges, on a dipole waveguide that's 12" wide by 6" tall. In a sealed box these woofers would play down to 100hz, on a dipole the F3 is just 800hz(!)

But the key to all of this is EQ; you can EQ the thing flat, and as long as you have enough power and enough excursion, you wind up with a loudspeaker that's loud on-axis, but quiet off-axis. Which means less interaction with the room.

 

[2010hummerguy]

Yep, EQ was key to my system as well. It actually sounded better if I played the OB15's up to 300hz and cut out the W22's completely. But I looked the way the W22's looked so I had a hard time getting rid of them

The sixe OB15's gave me enough excursion head room that heavy EQ was no big deal. I had smooth buttery flat output easy to 20hz...I needed to enact a subsonic filter just to keep the subs from ripping the baffles apart.

 

[Justin Zazzi]

Having listened to Linkwitz's LX521 dipole masterpieces at his house recently, I can confirm there is certainly something to this. He goes into great detail on his website about how the dipole system does not interact with the room as much as a normal (sealed) enclosure. The really neat thing about dipoles is they have almost zero radiation to the sides, so if you angle them properly, you can get zero first reflections off the side walls of your listening space without using acoustical treatments. Also, you can get away with using less bass traps because there is less bass energy going into the room. Also, with the radiation to the rear of the speakers, you get a sense of envelopment that is otherwise hard to come by (without the use of tons of diffusors).

Image from Linkwitz Labs here

It's a fascinating concept, and the demonstration was the most believable set of speakers I've heard anywhere at any price (including all the nutty stuff at CES and other super high-end shows). I bought a set of plans for the LX521 and hope to build them soon

 

[2010hummerguy]

Agreed, the reduced room interaction is amazing and the listener gets to benefit with a more realistic image.

The only system I prefer to a good dipole build is a well built line array but then you are multiplying the cost by a factor of +10x.

 

[JoshHefnerX]

Is the dipole giving better detail because of the control over the cone similar to an IB or is it really about the radiation pattern? I did notice going IB in my car that the detail was better than sealed or ported w/ the same sub.

Josh

 

[Orion525iT]

I have a dipole like arrangement in my coupe.

I basically stuffed eight 8" subs into the quarter panels. There are four subs in each quarter arranged in PPSL manifolds firing slightly forward into the cabin. The manifolds are as wide as possible, and the idea is to run the subs much higher into the midbass than what would be normal. Wide placement should improve stage width from greater ITD.

The rear wave from the manifolds vents into the quarters, over the wheel arch and into the trunk. I will not attempt to isolate the rear wave. However, it must travel the distance of the quarters and wheel arch all the way to the taillights before exiting into the trunk. The rear seat and absorbing materials/stuffing in the quarters will impact exactly where cancelation occurs, but it should be around the taillights area.

I don't know what you would call this arrangement, but I think it is essential a dipole.

In the experiments with manifolds I did in my living room, I found that manifolds helped improve bass output directly in front of the manifold. Output was greatly attenuated off axis.

I honestly have know idea what the final results of this arrangement will look like. I am not sure it has been tried before. I should be done with it soon, and will post measurements when I get a chance.

 

[Patrick Bateman]

Is the dipole giving better detail because of the control over the cone similar to an IB or is it really about the radiation pattern? I did notice going IB in my car that the detail was better than sealed or ported w/ the same sub.

Josh

There's simply less energy going into the room (or the car.)

It's similar to comparing a conventional lightbulb (omnidirectional) to a floodlight (directional)

It's pretty certain that waveguides get the job done better than dipoles or cardioids. But the problem with waveguides is that the size gets out of control in a hurry. For instance, to control directivity down to 500hz you need a waveguide that seven tenths of a meter wide(!) That's HUGE.

With a dipole or a cardioid, they can be small, in fact they work better if they're small. The downside is that the efficiency is TERRIBLE. But power is cheap, so that helps.

For me, the big "aha" moment was when I realized that if you have enough displacement, you can just brute force everything. My Dayton ND91s have an F3 of 800hz as a dipole, but you only need fifteen decibels of boost to make them flat to 200hz. That might sound like a lot of boost, but I'm using four of them. So I have twelve decibels more output than you'd get from a single driver.

And this efficiency has been verified; I'm applying FIFTEEN decibels of attenuation, just to get them to mate up with my tweeter! So basically I need a crap ton of EQ to get them to play down to 200hz, but I have a crap ton of headroom because I'm using so many midranges. (Four per side.)

 

[cajunner]

I like the Pluto.

 

[Patrick Bateman]

Agreed, the reduced room interaction is amazing and the listener gets to benefit with a more realistic image.

The only system I prefer to a good dipole build is a well built line array but then you are multiplying the cost by a factor of +10x.

They have some things in common too.

Here's an example:

1) If you have a 6.5" woofer in a sealed box, it's going to be omnidirectional from 2khz and below. Basically it's omnidirectional across most of it's bandwidth.
2) Put that same 6.5" woofer in a baffle that measures 6.5" x 6.5", and it's going to be dipole from 2khz and down. (A figure eight pattern.) So it's going to radiate to the front and to the back, but not to the left, right, up or down.
3) Stack ten of those 6.5" woofers in a vertical line of sealed boxes, and their horizontal directivity will be exactly the same as a single unit. (omni.) But now the vertical directivity is dramatically narrow. Basically the vertical radiation is narrow, down to 207hz.

That narrowing of vertical directivity negates the 'bounce' off the floor and the ceiling.

 

[knever3]

Dipole bass is my FAVORITE bass

Next dipole build will be with GR Research open baffle servo 12's in W frames. Should allow for narrower install and better response though the OB15's pictured above had no issue getting down. I miss that build but the next one will be even better

My boy would DESTROY that!!

Nice setup though.

 

[2010hummerguy]

My boy would DESTROY that!!

Nice setup though.

Yep it is all gone now, replaced by a more kid-friendly Rythmik FV15HP