OB MTM w/ Dipole Waveguide

[thadman]

With the recent popularity of open baffles and waveguides, I have been surprised to see little mention of combining the technologies.

If a waveguide was able to match the figure-eight response of the dipole, perhaps some interesting results could arise.

There are several candidates currently available which might be suitable for a dipole waveguide, specifically the B&G Neo3, ESS AMT, Beyma TPL150 AMT, among others.

An MTM might be able to provide the gradual transition necessary to match the horizontal directivity of the aforementioned tweeters.

Topics for discussion:

Crossover

Waveguide expansion (Le'Cleach, Min Phase, OS, etc)

Best Regards,
Thadman

[D OB G]

Thadman,

What would you think about forward and rearward facing waveguides with compression drivers, which would achieve a dipole pattern? A decision would need to be made as to whether to compromise the vertical coincidence by offsetting the magnets- diaphragms in the same plane, or a different sort of disruption to an ideal dipole pattern by having the magnets as close together as possible in the horizontal plane, front to back.

I do this with the magnets offset, and I roll off the response of the rear waveguide, about 3 dB per octave from about 4 kHz, where I believe the dipole pattern can't be maintained properly anyway at the wavelengths involved. (It sounds better this way!).

David

[thadman]

Originally Posted by D OB G

Thadman,

What would you think about forward and rearward facing waveguides with compression drivers, which would achieve a dipole pattern? A decision would need to be made as to whether to compromise the vertical coincidence by offsetting the magnets- diaphragms in the same plane, or a different sort of disruption to an ideal dipole pattern by having the magnets as close together as possible in the horizontal plane, front to back.

I do this with the magnets offset, and I roll off the response of the rear waveguide, about 3 dB per octave from about 4 kHz, where I believe the dipole pattern can't be maintained properly anyway at the wavelengths involved. (It sounds better this way!).

David

Interesting

Assuming your dipole waveguide transitioned to an open baffle alignment, how close were you able to match the response throughout the crossover? Were you able to achieve a similar figure-eight response or did it simply balance the reverberant field?

[LineSource]

Minimal size dipole baffles could have some advantages over waveguides for the mid-tweet. A dipole ribbon might add a minimal frame for diffraction reduction and limited directionality control. RAAL is recommending alternative bezel solutions for diffraction control.

If you have a digital crossover, it may be simplier to experiment with a minimal baffle mid-tweet before constructing larger waveguides that will better control but also greater limit dispersion angles. Dipole ribbons, planars, and Heil tweeters are getting a second look.

Goals? constant power at 45? 90? 120? 180? degrees?

[keyser]

I've been thinking along the same lines. As is said above, planar tweeters can get you almost there, potentially extending dipole behavior to about 6 khz. By adding a small wave-guided tweeter for frequencies above approximately 5 khz, or replacing the planar tweeter altogether, you might be able to extend the radiation-pattern even further.

However, physical alignment will be difficult and you'll probably get notable diffraction from the midrange driver against the wave-guide. Also, the center-to-center distance between mid and tweeter might become large, jeopardizing integration in the vertical plane.

If you can overcome these problems the idea could be very promising!

[Electron]

My implementation about the same theme is a bit different. It is Waveguide resistance box combination. Dipole does not work very close to the wall behind speaker but resistance loading (see the black areas on main unit sides) makes it to radiate only forward.
The picture below is not from the latest setup. Today the bass unit is different.
The main unit consist of 12 Hivi B3N and Infinity Emit II "ribbon". It is actually not pure line source because the acustical lenght of the line depends on the frequency so that on upper x-over range only two 3BN below tweeter are active and more and more comes in when frequency goes down.
radiation pattern to sides doesn not change with frequency. Sound in way just fades out.

Proto

[keyser]

Hi Electron, the picture is not visible if you don't have an account on that forum.

[D OB G]

Yes thadman, I'm really only balancing the reverberant field. I'm attempting to achieve the most even power response I can get.

I can't move the current speakers into the centre of the room to measure the rear polar response, but I can get an idea from previous iterations.

The rear radiation of the mid on the open baffle only produces a decent dipole up to about 500 Hz. By the 900 Hz of the crossover the back of the mid is getting a bit lumpy, even on axis (magnet?, although Linkwitz has an alternative view).

The dipole pattern actually improves again for a while where the tweeters do duty with the waveguides.

But in any case, mapping the rear radiation to the front, to determine the dipole pattern, is, IMHO, thwarted at higher frequencies by the room effects. I don't have a symmetrical room, nor are the speakers placed symmetrically in it. The furnishings behind and in front are different from side to side. This is not a dedicated HT room!!

And quite apart from the different absorptions at different frequencies at the rear of the speakers, and different reflected path lengths, I'm introducing a further complication by using a DEQX digital crossover, which time aligns at the FRONT, but makes things worse at the rear.

With diffraction issues, the waveguide is elliptical (18sound XT1086) and flush mounted hard up to the mid in the middle of the baffle. So I don't believe there are any more issues than the normal open baffle diffraction, TO THE FRONT.

Obviously I'm optimising for the front, pre-reflection response.

In summary I think a proper listening room might be necessary to obtain the true dipole benefits at higher frequencies (although constant directivity might only be realizable with waveguides i.e other methods may beam, as you imply), but the power response benefits apply in any case.

David

[Rudolf]

Originally Posted by keyser

However, physical alignment will be difficult and you'll probably get notable diffraction from the midrange driver against the wave-guide. Also, the center-to-center distance between mid and tweeter might become large, jeopardizing integration in the vertical plane.

Those above are only a few questions indeed

If we look at a dome tweeter like the DT-300 with WG it will measure like this:



This WG of 16 cm diameter will become truly CD only above 3 kHz. The gain of the WG (red line in second diagram) is barely utilized in that region.

If we mount two of these waveguides back to back on a baffle in the most narrow manner possible, we would need a baffle of 25 cm height, 16 cm width and 8 cm depth. This is already assuming the smallest imaginable magnet size of a neodymium tweeter instead of the big DT-300.
Simulating two such DT-300 in the appropriate positions on a baffle of above dimensions will result in this horizontal response:



The simulation program does not account for the WGs off-axis-rolloff, so you have to imagine the 30° and 60° line somewhat lower in SPL. But it becomes obvious how the sheer size of the WGs will destroy any CD at 2-5 kHz.

If you can overcome these problems the idea could be very promising!

I believe that for any given frequency a waveguide with true constant directivity must be too large in size to allow for an appropriate dipole length (distance of the two dipole sources) at that frequency. CD waveguides that work in a truely dipole fashion are physically impossible IMHO.

Rudolf

[markus76]

Why is nobody evaluating the dipole proposal Earl made? He mentioned putting the driver in a donut shaped form several times.