I'm curious:
I've been reading a ton about open baffle designs, ranging from huge baffles to fully "nude" systems, and in all combinations of dipole + monopole combinations across the frequency spectrum.
Since most tweeters are monopole, many designers add tweeters on the back of the baffle/system. I've even seen this on closed box systems. It made me wonder in which ways that kind of solution might be better, different, or worse than dipole tweeters?
Could this extend to the midrange, say, with 2 mids mounted back to back in a tube the diameter of the drivers, wired out of phase?
What would happen as the wavelengths get longer... maybe identical sealed (or ported) chambers front and back, with the drivers again wired out of phase?
I'm curious whether any advantages (or disadvantages) could ensue:
- How many ways are there to generate dipole patterns?
- Do any of the alternate approaches have particular advantages and disadvantages to learn from?
I've been reading a ton about open baffle designs, ranging from huge baffles to fully "nude" systems, and in all combinations of dipole + monopole combinations across the frequency spectrum.
Since most tweeters are monopole, many designers add tweeters on the back of the baffle/system. I've even seen this on closed box systems. It made me wonder in which ways that kind of solution might be better, different, or worse than dipole tweeters?
Could this extend to the midrange, say, with 2 mids mounted back to back in a tube the diameter of the drivers, wired out of phase?
What would happen as the wavelengths get longer... maybe identical sealed (or ported) chambers front and back, with the drivers again wired out of phase?
I'm curious whether any advantages (or disadvantages) could ensue:
- Dipole sound, "closed box" aesthetic?
- Instantly change your speaker from being dipole to monopole depending on the music or mood or room placement (e.g. shoved to wall or not) using DSP?
- Other thoughts?
One opinion:
https://www.linkwitzlab.com/models.htm
There is not much sense in building a dipole speaker with two small closed box speakers driven in opposite phase, when one of the objectives is to remove the sound character that boxes impart. The two boxes a can be joined at their backs, though, and the connecting wall removed
. . .
The closed dipole baffle b with two drivers can be evolved d into the flat, circular, open baffle e with a single driver, while maintaining the same excursion limited output capability.
The circular baffle with a -|+ point source at its center has nearly the same polar radiation pattern as the two opposite polarity point sources spaced D apart in the model of figure A1 above. The circular baffle's usefulness at high frequencies is limited by the sharp interference nulls when D is a multiple of a wavelength. This behavior can be considerably smoothed by making the baffle f rectangular which gives a variation to the length of D.
https://www.linkwitzlab.com/models.htm
There is not much sense in building a dipole speaker with two small closed box speakers driven in opposite phase, when one of the objectives is to remove the sound character that boxes impart. The two boxes a can be joined at their backs, though, and the connecting wall removed
. . .
The closed dipole baffle b with two drivers can be evolved d into the flat, circular, open baffle e with a single driver, while maintaining the same excursion limited output capability.
The circular baffle with a -|+ point source at its center has nearly the same polar radiation pattern as the two opposite polarity point sources spaced D apart in the model of figure A1 above. The circular baffle's usefulness at high frequencies is limited by the sharp interference nulls when D is a multiple of a wavelength. This behavior can be considerably smoothed by making the baffle f rectangular which gives a variation to the length of D.
I used to run two pairs of small English minimonitors, biamped with identical amps.
One on top of the other, the top one upside down and facing the back wall. I could easily make them run in phase ( bipole ) or out of phase ( dipole ) by swapping the cables on the top one.
Also, if I felt like it, I could rotate the top one facing forwards too in a pseudo d'Appolitto configuration ( and 3db louder! ). This was the normal, "at rest" configuration so I didn't have to show off the wiring when not in use ( I had them biwired too... ).
Sure, given they were minimonitors, baffle size and bass were not a big issue and I drove a pair of servo woofers separately.
It could be ugly to look at, but I loved them and it gave me a big lesson on sound.
One on top of the other, the top one upside down and facing the back wall. I could easily make them run in phase ( bipole ) or out of phase ( dipole ) by swapping the cables on the top one.
Also, if I felt like it, I could rotate the top one facing forwards too in a pseudo d'Appolitto configuration ( and 3db louder! ). This was the normal, "at rest" configuration so I didn't have to show off the wiring when not in use ( I had them biwired too... ).
Sure, given they were minimonitors, baffle size and bass were not a big issue and I drove a pair of servo woofers separately.
It could be ugly to look at, but I loved them and it gave me a big lesson on sound.
Last edited:
I chose to use dual amps and speakers because I had them around the I wanted to experiment.
Facing them forward in a pseudo d'Appolito configuration affected the imaging, but just a very little bit since the mini monitors have imaging in spades, but helped out the dynamics and extended the bass. It was a worthwhile test. So, the search for the bigger brothers started ( the speakers were Acoustic Energy AE1, so the search was for the AE2).
In the meantime, the chance was there to play around, so off we went. I didn't so much care about the bipole configuration, but I did like the dipole set up... lots of spaciousness while still keep the pin point accuracy of the speakers...
In fact, I liked the dipole set up so much that I went and got Maggies and never got the AE2.
Facing them forward in a pseudo d'Appolito configuration affected the imaging, but just a very little bit since the mini monitors have imaging in spades, but helped out the dynamics and extended the bass. It was a worthwhile test. So, the search for the bigger brothers started ( the speakers were Acoustic Energy AE1, so the search was for the AE2).
In the meantime, the chance was there to play around, so off we went. I didn't so much care about the bipole configuration, but I did like the dipole set up... lots of spaciousness while still keep the pin point accuracy of the speakers...
In fact, I liked the dipole set up so much that I went and got Maggies and never got the AE2.
That's a bit surprising because the narrow directivity of full planars like Magneplanars & electrostatics sound so different to me from dynamic driver OB dipoles which generally have much wider dispersion both front & back. Sweet spot positioning isn't critical with my Linkwitz-inspired dipoles. Not so, in my experience, with Martin-Logan electrostatics or big Maggies.
Maggies do have a sweet spot.... It's sort of different from point drivers... you move them sideways until they hit the right spot... it's not just a matter of rotating them until they face you, or in or out... With Maggies I strongly recommend you get a big Persian rug and remember the positioning of the speakers on the rug pattern. Hmm.. DON'T ever move the rug!
The AE1 "dipoles" also had a sweet spot.
Others put tape on the floor..
Now, spaciousness.. that's a different matter and this is one of the things I love about a dipole set up. That's why I went with dipoles, I like that combination of spaciousness with tight imaging. Instruments really hang in their own space.
The AE1 "dipoles" also had a sweet spot.
Others put tape on the floor..
Now, spaciousness.. that's a different matter and this is one of the things I love about a dipole set up. That's why I went with dipoles, I like that combination of spaciousness with tight imaging. Instruments really hang in their own space.
I have built three speakers that can play as a dipole or bipolar speaker depending on how the drivers are connected. I spent a lot of time modeling the speakers in Visaton boxsim to see how the directivity / radiation pattern changed with baffle thickness and size. That software lets you place drivers on any surface of a box, front/rear, sides. I first built a pair of 6" diameter spheres with 20 one inch drivers. I grouped them front cap, rear cap, ring around the middle. So with three amp channels and DSP I could configure them dipole, omni, or point source using delays. A mix of the omni with some delay was preferred by most people that listened to them. As they sold after two weeks I built a pair of 4" diameter spheres using just six 2" drivers with similar capability. The most recent build is a three way with the mid and tweet drivers mounted back to back. This allows for dipole or bipolar (omni directional) patterns. I selected 3" midranges with a shallow cup and set crossover frequencies low, 425Hz and 2,300 Hz to achieve a near omni pattern full range. To improve the response I need to make the midrange / tweeter top baffle much wider to lower the baffle step frequency.
Driver configuration Visaton Boxsim
Directivity calculated with Visaton Boxsim freeware
Driver configuration Visaton Boxsim
Directivity calculated with Visaton Boxsim freeware
Attachments
Last edited:
Last year I measured and then retrofitted a pair of 3.7i Maggies with digital crossovers for a friend. The passive crossovers had shallow and variable slopes such that the drivers overlapped in response. So the off axis response horizontally had huge variations. With the ability to EQ the individual drivers and implement higher order crossovers the 3.7i was transformed into a very accurate speaker with a much wider sweet spot.
What seriously awesome experiments! The first 2 redefine the concept of "satellites"! Plus points for the dedication to wire up 20 of the 1" drivers per side.
How different did the 6x 2" drivers sound? How loud and low could they play?
For the three-way design, I'm curious why you skipped the bass section?
Thanks so much for sharing!
Not shown were two 12" woofer boxes that operated with the spheres. I documented the spheres pretty well in this thread. https://www.diyaudio.com/community/threads/how-to-build-a-spherical-speaker.374278/post-7418913
There were several differences between the four and six inch spheres.
1. 40 drivers are a lot more money than 12 drivers.
2. A pulsing sphere has a high frequency cutoff of sorts that depends on the diameter. The sound from the sides comes around with delay and adds to the sound from the front, but this happens in every direction. So the 6" sphere had the highs roll off at a lower frequency than the 4" sphere when operated as a full omni, with all drivers playing the same content.
3. How loud. The 20 driver sphere 111.5 dB and the 6 driver sphere 109 dB SPL at 300Hz, at Xmax, at one meter. So the 20 driver spheres exceed what most home speakers will do. For comparison the seven foot tall Infinity IRSV line array also has 20 tweeters.
So the 20 drivers play cleanly at high levels where the smaller sphere run out of gas. Due to the larger sphere more high frequency EQ was needed. As I have never heard them side by side, it's my best recollection. The omni needs a higher SPL to sound as loud as a forward firing speaker as the sound is distributed in all directions and it's very dependent on the room and how many people ( mobile sound absorbers ) are there.
Here's a link to my SPL calculation spreadsheet. The Aurasound drivers are in there.
https://docs.google.com/spreadsheets/d/18BXQP-WSRRbQHTR_mXJ0OVdyYF8tixPRP3dvkLzf5m0/edit?usp=sharing
"For the three-way design, I'm curious why you skipped the bass section?"
I didn't. It's just not obvious. The three way design uses two bass boxes per side, each with a 10" woofer and a 12" passive radiator. As I can't lift heavy objects, I made these boxes as light as possible, just 20 lbs each, and stack them. To get the desired effect one woofer facing forward and the other woofer facing rearward. As they are stacked, the woofers are not back to back, but at low frequencies the vertical offset does not seem to matter. With this configuration the speaker can be carried as four manageable parts, two woofer boxes, the mid/tweet box and the crossover.
I posted a few times about the three way on the speaker contest thread. https://www.diyaudio.com/community/...-speaker-contest-for-2024.416557/post-7808070
There were several differences between the four and six inch spheres.
1. 40 drivers are a lot more money than 12 drivers.
2. A pulsing sphere has a high frequency cutoff of sorts that depends on the diameter. The sound from the sides comes around with delay and adds to the sound from the front, but this happens in every direction. So the 6" sphere had the highs roll off at a lower frequency than the 4" sphere when operated as a full omni, with all drivers playing the same content.
3. How loud. The 20 driver sphere 111.5 dB and the 6 driver sphere 109 dB SPL at 300Hz, at Xmax, at one meter. So the 20 driver spheres exceed what most home speakers will do. For comparison the seven foot tall Infinity IRSV line array also has 20 tweeters.
So the 20 drivers play cleanly at high levels where the smaller sphere run out of gas. Due to the larger sphere more high frequency EQ was needed. As I have never heard them side by side, it's my best recollection. The omni needs a higher SPL to sound as loud as a forward firing speaker as the sound is distributed in all directions and it's very dependent on the room and how many people ( mobile sound absorbers ) are there.
Here's a link to my SPL calculation spreadsheet. The Aurasound drivers are in there.
https://docs.google.com/spreadsheets/d/18BXQP-WSRRbQHTR_mXJ0OVdyYF8tixPRP3dvkLzf5m0/edit?usp=sharing
"For the three-way design, I'm curious why you skipped the bass section?"
I didn't. It's just not obvious. The three way design uses two bass boxes per side, each with a 10" woofer and a 12" passive radiator. As I can't lift heavy objects, I made these boxes as light as possible, just 20 lbs each, and stack them. To get the desired effect one woofer facing forward and the other woofer facing rearward. As they are stacked, the woofers are not back to back, but at low frequencies the vertical offset does not seem to matter. With this configuration the speaker can be carried as four manageable parts, two woofer boxes, the mid/tweet box and the crossover.
I posted a few times about the three way on the speaker contest thread. https://www.diyaudio.com/community/...-speaker-contest-for-2024.416557/post-7808070
Last edited:
Tre real challenge is how to get uniform dipole pattern throughout the whole range. Dipole axial nulling has omnipolar pattern and it makes noticable coloration in-room. Below 2-3kHz is easy but above short wavelength makes serious challenges.
Narrow/no baffle helps but smallest cone or planar drivers are 5-6cm wide - nulling around 6-7kHz and might start beaming before that. That is already at the limit of pure music notes, territory of harmonics with low spl so less important, question is about "airiness" etc. hifi jargon.
Double drivers in opposing polarity is waste of resources, but ok above 6 kHz to add "airiness" which can be measured in room decay.
Narrow/no baffle helps but smallest cone or planar drivers are 5-6cm wide - nulling around 6-7kHz and might start beaming before that. That is already at the limit of pure music notes, territory of harmonics with low spl so less important, question is about "airiness" etc. hifi jargon.
Double drivers in opposing polarity is waste of resources, but ok above 6 kHz to add "airiness" which can be measured in room decay.
For a pure dipole build that is most often true. Floor space is a precious resource. The cost of the double drivers provides the benefit of one speaker that can easily be switched between multiple radiation patterns on the fly often with a remote control. So that becomes important when you start running out of space in your room for more and more speakers. I'm sure no one has that problem. Ha
Not only that, most such arrangements will lead to an increased distance between the two dipole sources, which is going to broaden the radiation pattern as frequency rises.
Websites about dipole design and speakers
https://gainphile.blogspot.com/
http://www.dipolplus.de/
https://www.linkwitzlab.com/models.htm
https://musicanddesign.speakerdesign.net/naomain.html
https://gainphile.blogspot.com/
http://www.dipolplus.de/
https://www.linkwitzlab.com/models.htm
https://musicanddesign.speakerdesign.net/naomain.html
Might it also increase the range of usable drivers? For example, I'm recalling notes from @CharlieLaub about the rear side radiation measurements often being significantly different from the front side. Does the efficiency change as well?
Returning to the original idea of this thread, I'm curious what other approaches there are and respective pros/cons. I'm definitely not arguing for any particular viewpoint. 🙂
In my experiments with boxsim, I found that the delay due to increased distance between the two sources resulted in frequency response disturbances below the frequency where the driver directivity is greater than 180 degrees and the baffle is narrow enough to allow that radiation to come around to interfere with the other driver.
It seems from your comment that you are counting on this low delay wrap around radiation to cause cancelation to narrow the directivity pattern. I will have to look at that with a simulation. If I was going for a narrow forward pattern speaker I would lean towards a cardioid pattern speaker with directivity controlled with waveguides and cone diameter as well as cancelation. I always figured the interesting quality of an open baffle speaker was it's ability to direct sound in a wide pattern. Limiting it to only having a back wall reflection with a narrow pattern might be interesting. It would seem that's close to the concept the Bose 901 was built around.
What about a cardioid pattern? I haven't tried it, but it could be interesting if trying to reduce reflections and room loading at low-ish frequencies, while the higher frequencies could optionally provide more air. At least that's how I imagine it.
Without overthinking it: 2 channels per side and active (all-pass) filters that provide a 90°-180° phase shift for the desired range.
Without overthinking it: 2 channels per side and active (all-pass) filters that provide a 90°-180° phase shift for the desired range.
Not fair... you went down to the Costco Kirkland Factory where they make toilet paper and wine, raise chickens and design the coolest spherical speakers...
You sold them....
I've been looking at DSP crossovers... many of them digitize the signal to 48Khz... down rez'd. A very few run 24/96. Ideally, we should get a DSP that has a digital signal and does not include a DAC... a little box that connects via USB in and out. Do not resample the signal.
It could be just a software program too...