dipoles

[exipnos]

Hi there,

Continuing on my quest to learn more about dipoles and HT. A common recommendation is to use dipoles for HT in the surround channels especially back and sides. I've seen that all the commercial dipoles for HT are boxed designs with 2 sets of drivers shooting in opposite directions. The placement of these dipoles are on the wall.

By doing OB dipoles we remove the box and its effect on sound. For surrounds, would placement of OB bipoles be the same as closed boxes? I imagine placing the speaker parallel to the side wall so that the front wave shoots to the front of the HT and back wave goes to the back.

Closed boxes have 2 sets of tweeters. For OB dipoles I imagine that I would have to use a tweeter where the back wave is used. That means some sort of ribbon or planar. I 'm thinking of using a Swedish design for a diy ribbon (http://www.dahlbergaudiodesign.se/). Any suggestions for a commercial one, prefeerable cheap?

[Victor]

I have a dipole-based HT. That is to say I have a 20-Hz capable dipole speaker in each corner of my HT. I can achieve ear-bleeding volume levels at sub 1% distortion at frequencies above 50 Hz. My take on dipoles is that it is certainly a way to go. You, however, need a fairly large room.

Dipole speaker must be placed at least 1-2 feet from any wall for optimum performance. However, in practice I noticed that the side of the speaker can be placed next to the wall, but the back must be at least 1 foot away from the back wall, with 2 feet giving somewhat better results.

Dipole speakers capable of high SPL are large, requiring multiple drivers. The dipole subwoofer in reality does not exist, - that is to say you cannot achieve 20-Hz performance at 100 dB of more necessary for modern DVD sound tracks. You can do about 92-96 with multiple ( at least 4) 12-inch drivers. I found 15-inch drivers to noisy for dipole application, although I know many people use them. Best scenario is to build a sealed sub with 15 or even 18-inch drivers for below 40 Hz duty, with IB being the ultimate.

My preference is also for line-arrays. They work very well in a dipole. I use RD-75 in the front and Carver planar magnetic in the rear. Dipoles will need active equalization, although passive approach is possible but it is far from trivial. Active EQ and active cross-over is a way to go. If you use analog circuits, by the time you are done you’ll have 2 or more dozens op-amps pressed into service. I use a fully digital signal processing chain and it works well. If you are careful about choosing your drivers then you may get away with DSP-based crossover boxes. The signal headroom in a digital cross-over is a limiting factor here. If your drivers do not require more then 10-12 dB of EQ then DSP approach is viable.

You must also think about the Volume Control before you start. With analog EQ and cross-over the Volume Control implementation is simple, as you can use a 2-channel high quality potentiometer. With a digital cross-over, it gets complicated. I use Roland M-1000 digital mixers for this purpose. Some will argue that I loose resolution, because the M-1000 must drop bits to attenuate. Well, I do not for 16-bit sources, but I will loose 1-2 bits at most for 24 bit sources when they appear. When one considers that the best S/N ratio for any source regardless of the price is about 100 dB at the analog output, the loss of 1-2 bits of resolution is irrelevant.

Finally, dipoles may get very expensive and WAF can be critically low.

Well, in a nutshell this is my take on dipoles, - Good Luck
Victor.

[ThomasW]

Sorry to say that "cheap" and "ribbon" are mutually exclusive terms.

Though not technically ribbons, (they're planar magnetic) the B&G drivers like Victor and I use, are some of the lowest cost 'better' dipole planar units available.

Something else to consider is either the Magnepan MMG or MMGW. The MMG's only need a sub, the MMGW benefit from a midwoofer. http://www.magnepan.com/

[exipnos]

I
Dipole speaker must be placed at least 1-2 feet from any wall for optimum performance. However, in practice I noticed that the side of the speaker can be placed next to the wall, but the back must be at least 1 foot away from the back wall, with 2 feet giving somewhat better results.

Thanks for your take on dipoles. It helps.

So for placement I guess I should be fine if I can place it next to the wall. This is what I had in mind.

1. Side surrounds. Directly against the wall facing front and back to the room. Ample space both in front and behind the speaker. Should be okay right?

2. Back surrounds. Directly aginst the wall but drivers facing left and right. About 3-4 ft space to the side wall. I guess this should also be okay?

3. Front surrounds and center. Not sure if I should use dipoles here. Wouldn't you want direct radiating speakers here? But on the other hand most people use dipoles for the front speakers so maybe this is the way forward. I guess importance of placement here is to placed a couple ft from the wall. Seating and speaker location should be placed so that the front wave reflects from the back of the room and hits the seat in the same time as the back wave have reflected from the front wall and returns to the seat. Is that correct?

[Landroval]

1. Side surrounds. Directly against the wall facing front and back to the room. Ample space both in front and behind the speaker. Should be okay right?

2. Back surrounds. Directly aginst the wall but drivers facing left and right. About 3-4 ft space to the side wall. I guess this should also be okay?

Uuh, neither of those will work. Your listening postion will probably will be quite directly at the side of the dipole where the dipole's front and back waves are nullifying each other. You'll hear something between nothing and weird out of phase sound.

The boxed "dipole" surround speakers are actually bipoles (all drivers in the same phase) and/or have a bipole/dipole/monopole radiating switch. Normally they're used as bipoles, and I believe the dipole option is used only in situations where the speaker is close to a corner and the nearby wall is causing the unwanted phase change (~or something like that).

I'm not saying your design is impossible to make work, so if you feel like it, you could try it and try to find the best place and orientation for the speakers. Maybe it could come to something.

[exipnos]

Oh no! Now I'm even more confused then before. Does this comment come from your experience?

Yes the listener will probably be placed in the null. But accordind to most documentation I read this is desired. Dipoles in HT are supposed to provide a difused soundfield. I'm not looking for speakers that will be used for multichannel music. I belive there you would like monopole sound.

I've seen that most dipoles have switches for dipole, bipole, monopole and sometimes tripole operation. But my understanding was that dipole was mainly used.

I'm not sure what to say. There is no rush for me to decide the speaker design since construction won't start until this spring. I would welcome more comments on this.

[Landroval]

Hmm, I might have been a little wrong here. The dipole soundfield might be just the right thing here. Although most commercial designs are not 'pure' dipoles because they have a monopole bass to help with decent bass output. Nor are the front and back soundfields on the same axis, but more or less turned towards the listener to make the null area smaller.