I am planning a sort of "ultimate" dipole loudspeaker based on some recent projects. This is a 3-way design that is typically augmented by stereo subwoofers for the lowest octaves. The driver(s) in the woofer section are mounted in a relatively small (approximately 20x20 inches or 0.5x0.5m) baffle that sits on the floor. I currently have a speaker of this design that uses four 8" woofers that are pretty decent. For the "ultimate" version I want to kick it up a notch, and I think I have driver options narrowed down to these three. I am trying to consider the cost and performance of each and would like some input. I have ruled out using drivers larger than 12" for reasons that I will not get into now, and this choice had led me to the following possibilities:
Acoustic Elegance LO12:
This driver costs about $440 each. AE has a very good reputation for drivers based around the Lambda motor, and notably for OB use. The LO series has a little higher sensitivity and lower Xmax than e.g. the dipole line, but supposedly has a cleaner upper end to the passband. This is important in my application due to baffle losses tilting the response curve and how that impacts the relative levels of distortion. I would use a single driver per side. The low end is restricted by total displacement and for this driver Sd = 530cm^2 and Xmax = 9mm (mechanical limit +/- 12mm). I have seen some old measurements of the LO15 and these look encouraging in terms of distortion. Also important for judging power input is Q=0.66 and Fs=42Hz and 2.83V sensitivity is 97dB. This could work very well indeed.
Scan-Speak 32W/4878T11:
This is a higher sensitivity, lower Xmax version of the 32W/4878T00 subwoofer that was released a few years ago. The T00 motor and suspension are very linear, and I assume that the T11 is similar but with some modifications for higher sensitivity. Relevant parameters are: Sd = 526cm^2, Xmax = 7mm (mechanical limit of +/- 28mm), Qts = 0.21, Fs = 24Hz, 2.83V sensitivity = 95dB. These are more expensive at about $675 each, so more expensive than the LO12 but may be slightly better in terms of distortion performance. The low Q may require about 6dB more power input at the low end of the passband. Otherwise this is very high performance option but costs about 50% more.
Four Purifi PTT8.0X08:
This recently released driver from Purifi has a very high level of performance in areas that the 12" drivers do not. Because four (per side) would be used, total Sd is about double of a single 12" driver. This may mean that subwoofers are not necessary. Distortion performance, both HD and IMD, is exceptional. The downside is that these drivers are very expensive, each costing about the same as one 12" 32W/4878T11. I don't even want to write that number down - it's so large. As I mentioned this will be an "ultimate" speaker, and paying the ultimate price for ultimate performance might just be justified. It's a tough call. Using four 8" drivers makes the construction of the speaker a little more challenging, but the concept has been proven in my current prototype. Relevant parameters are (for four drivers combined, where appropriate) Sd = 940cm^2 total, fro each driver Xmax = 9mm (mechanical limit of +/- 15mm), Qts = 0.35, Fs = 24Hz, 2.83V sensitivity = 92dB (series parallel connection) or 95dB (2x 2 in parallel, each pair driven by a separate amplifier channel). Sensitivity is raised via the electrical connection, and Sd is about double a single 12" so the Xmax value of 9mm is more like a 18-19mm Xmax for a single 12" driver. Another way to look at it is that the total Sd is about 15% larger than a single 15" driver having the same Xmax. The additional total displacement and very low distortion performance should permit operation to about one octave lower than a single 12", and this is low enough that a subwoofer is not needed since the woofers can probably operate down to the mid 30 Hz area. Without a pair of boxed subs, the whole loudspeaker system is smaller and easier to set up, transport, etc.
I have to say that I am very lucky that I can even consider any of these drivers. I'd like to get some feedback on these choices from readers who might be familiar with at least the first two drivers listed. I don't want to say much about the rest of the loudspeaker design, since details will come out (hopefully soon) via an article in audioXpress that is currently awaiting publication.
Acoustic Elegance LO12:
This driver costs about $440 each. AE has a very good reputation for drivers based around the Lambda motor, and notably for OB use. The LO series has a little higher sensitivity and lower Xmax than e.g. the dipole line, but supposedly has a cleaner upper end to the passband. This is important in my application due to baffle losses tilting the response curve and how that impacts the relative levels of distortion. I would use a single driver per side. The low end is restricted by total displacement and for this driver Sd = 530cm^2 and Xmax = 9mm (mechanical limit +/- 12mm). I have seen some old measurements of the LO15 and these look encouraging in terms of distortion. Also important for judging power input is Q=0.66 and Fs=42Hz and 2.83V sensitivity is 97dB. This could work very well indeed.
Scan-Speak 32W/4878T11:
This is a higher sensitivity, lower Xmax version of the 32W/4878T00 subwoofer that was released a few years ago. The T00 motor and suspension are very linear, and I assume that the T11 is similar but with some modifications for higher sensitivity. Relevant parameters are: Sd = 526cm^2, Xmax = 7mm (mechanical limit of +/- 28mm), Qts = 0.21, Fs = 24Hz, 2.83V sensitivity = 95dB. These are more expensive at about $675 each, so more expensive than the LO12 but may be slightly better in terms of distortion performance. The low Q may require about 6dB more power input at the low end of the passband. Otherwise this is very high performance option but costs about 50% more.
Four Purifi PTT8.0X08:
This recently released driver from Purifi has a very high level of performance in areas that the 12" drivers do not. Because four (per side) would be used, total Sd is about double of a single 12" driver. This may mean that subwoofers are not necessary. Distortion performance, both HD and IMD, is exceptional. The downside is that these drivers are very expensive, each costing about the same as one 12" 32W/4878T11. I don't even want to write that number down - it's so large. As I mentioned this will be an "ultimate" speaker, and paying the ultimate price for ultimate performance might just be justified. It's a tough call. Using four 8" drivers makes the construction of the speaker a little more challenging, but the concept has been proven in my current prototype. Relevant parameters are (for four drivers combined, where appropriate) Sd = 940cm^2 total, fro each driver Xmax = 9mm (mechanical limit of +/- 15mm), Qts = 0.35, Fs = 24Hz, 2.83V sensitivity = 92dB (series parallel connection) or 95dB (2x 2 in parallel, each pair driven by a separate amplifier channel). Sensitivity is raised via the electrical connection, and Sd is about double a single 12" so the Xmax value of 9mm is more like a 18-19mm Xmax for a single 12" driver. Another way to look at it is that the total Sd is about 15% larger than a single 15" driver having the same Xmax. The additional total displacement and very low distortion performance should permit operation to about one octave lower than a single 12", and this is low enough that a subwoofer is not needed since the woofers can probably operate down to the mid 30 Hz area. Without a pair of boxed subs, the whole loudspeaker system is smaller and easier to set up, transport, etc.
I have to say that I am very lucky that I can even consider any of these drivers. I'd like to get some feedback on these choices from readers who might be familiar with at least the first two drivers listed. I don't want to say much about the rest of the loudspeaker design, since details will come out (hopefully soon) via an article in audioXpress that is currently awaiting publication.
Last edited:
Hi Charlie,
You´re offering 2 different concepts, with and without subwoofer... and there are alternatives for each case. First with subwoofer calls for the best 12" out there, but a little bit bigger at 15" will give you a bit more meat on the bones to meet your subwoofer, and many 15" go high enough clean to meet your mids. I´d have a look at FC solutions from EMS, Supravox, Kanzen or PM alternatives from AE, Sonido (only 12" available, but maybe custom 15" is possible), Fostex or JBL... there are a lot of drivers out there.
Second concept without subwoofer, to have realistic bass from a small baffle calls for a lot of air to be cleanly moved, at least 2 12", better 1 18", but of course 4 8" are a possibility, too.
Personally, I´m still happy with AE Dipole 18 (despite it´s faults, namely too high a Mms) and haven´t found feasable alternatives. Even with its high Mms of nearly 200 gr Dipole 18 is fine between 25 and 350 Hz, but needs a lot of power.
With a 12", you´re dealing with, lets say 50 gr of Mms, and your baffle has to cater for that, to withstand and suppress induced vibration. That´s feasable, but with a 15" or even 18" intended to make bass on a smallish baffle, you´re easily in MMs regins of 100-200 gr, and the driver will shake your baffle...
What are your plans for mid/high?
All the best
Mattes
You´re offering 2 different concepts, with and without subwoofer... and there are alternatives for each case. First with subwoofer calls for the best 12" out there, but a little bit bigger at 15" will give you a bit more meat on the bones to meet your subwoofer, and many 15" go high enough clean to meet your mids. I´d have a look at FC solutions from EMS, Supravox, Kanzen or PM alternatives from AE, Sonido (only 12" available, but maybe custom 15" is possible), Fostex or JBL... there are a lot of drivers out there.
Second concept without subwoofer, to have realistic bass from a small baffle calls for a lot of air to be cleanly moved, at least 2 12", better 1 18", but of course 4 8" are a possibility, too.
Personally, I´m still happy with AE Dipole 18 (despite it´s faults, namely too high a Mms) and haven´t found feasable alternatives. Even with its high Mms of nearly 200 gr Dipole 18 is fine between 25 and 350 Hz, but needs a lot of power.
With a 12", you´re dealing with, lets say 50 gr of Mms, and your baffle has to cater for that, to withstand and suppress induced vibration. That´s feasable, but with a 15" or even 18" intended to make bass on a smallish baffle, you´re easily in MMs regins of 100-200 gr, and the driver will shake your baffle...
What are your plans for mid/high?
All the best
Mattes
I tried a number of 15, 18, and 21" pro drivers last year and I was not satisfied with most. By "try" I mean buy, and that gets expensive. The only one I found to work OK was a JBL 2268H, and I am using that in a system now. That driver is used nude, without a baffle, but being an 18 that is essentially the baffle size. So I want to keep the driver at 12" or less and put that in a baffle.
The vibration issue is not all that bad in my design. The woofer "panel" does not contain any other drivers and sits on the floor, so it is relatively stable. I cross over to subs around 70-80Hz so the driver excursion is not extreme nor is panel movement. The other drivers are 100% mechanically isolated from the woofers.
The vibration issue is not all that bad in my design. The woofer "panel" does not contain any other drivers and sits on the floor, so it is relatively stable. I cross over to subs around 70-80Hz so the driver excursion is not extreme nor is panel movement. The other drivers are 100% mechanically isolated from the woofers.
Another driver could be added to the list above: the Seas L26RO4Y.
These drivers are used in the Linkwitz LX521 in an M frame type of enclosure. I can say based on my own listening experience with the LX521 that these drivers are very clean at high excursion. During the development, several aspects of the driver were optimized for dipole use to meet Siegfried Linkwitz's goals for his system. Using these drivers might enable a two different variations: use 2-per side and augment their output with a subwoofer below 50Hz when needed, or use 4-per-side for 100% dipole operation. Relevant parameters are: Sd = 342cm^2, Xmax = 15mm (mechanical limit of +/- 28mm), Qts = 0.27, Fs = 26Hz, 2.83V sensitivity = 86dB (but that is being a little generous with the SPL). They are also more of a subwoofer type of drivers I guess, so the low sensitivity is not too unexpected. On the plus side they are not super expensive, at about $400 each, and a little less in quantity.
I recently read about a loudspeaker from the company Steinway Lyngdorf, the Model B. This is a large, heavy, and extremely expensive dipole loudspeaker system. It uses six 12" underhung woofers per side. This video of the company owner describing the speaker and its properties aligns very much with my own perspective on dipole systems:
The problem using so many woofer is that they occupy quite a lot of real estate. I prefer my speakers to be more compact and have a small visual footprint, so that sort of approach would not align with my goals.
These drivers are used in the Linkwitz LX521 in an M frame type of enclosure. I can say based on my own listening experience with the LX521 that these drivers are very clean at high excursion. During the development, several aspects of the driver were optimized for dipole use to meet Siegfried Linkwitz's goals for his system. Using these drivers might enable a two different variations: use 2-per side and augment their output with a subwoofer below 50Hz when needed, or use 4-per-side for 100% dipole operation. Relevant parameters are: Sd = 342cm^2, Xmax = 15mm (mechanical limit of +/- 28mm), Qts = 0.27, Fs = 26Hz, 2.83V sensitivity = 86dB (but that is being a little generous with the SPL). They are also more of a subwoofer type of drivers I guess, so the low sensitivity is not too unexpected. On the plus side they are not super expensive, at about $400 each, and a little less in quantity.
I recently read about a loudspeaker from the company Steinway Lyngdorf, the Model B. This is a large, heavy, and extremely expensive dipole loudspeaker system. It uses six 12" underhung woofers per side. This video of the company owner describing the speaker and its properties aligns very much with my own perspective on dipole systems:
The problem using so many woofer is that they occupy quite a lot of real estate. I prefer my speakers to be more compact and have a small visual footprint, so that sort of approach would not align with my goals.
If I were in your shoes, I would opt for dipole until some 80 Hz and then an optimally placed set of closed box dedicated subwoofers for the 30-80Hz range.
That is what I am doing now with a couple of prototype systems. Dipoles just don't do all that well below 30Hz, and if you choose 70 or 80 Hz as the low end cutoff for the dipoles the displacement requirements are much lower. But a subwoofer represents another band, which requires more amplifier channels and boxes. So if I am OK with great dipole response down to 30 Hz and am willing to forego a subwoofer I would need several low distortion drivers to pull that off since I do not want to use much baffle area.
Well, a decent dipole suitable sub in the category you mentioned will easily cost you way more than an amp with some power. You will be surprised how little power an unequalized subwoofer consumes. But that's my take, certainly if baffle area/footprint is a serious consideration.
I believe Purifi have plans to bring out a 10“; I’m guessing a couple of those per side would probably meet your needs.
Lars Risbo posted in the thread about XRK’s XSD speaker that one of his partners (I forget which one but not Bruno) likes using the Purifi drivers in OB . . . .
Lars Risbo posted in the thread about XRK’s XSD speaker that one of his partners (I forget which one but not Bruno) likes using the Purifi drivers in OB . . . .
That's right. Lars told me that the new 10" model won't be available to purchase until Fall 2023 (at least). Full TS parameters are not yet released but IIRC 15mm Xmax.
Thanks for posting that video. I wasn't aware of that model B. Very knowledgeable and enthusiastic man. The bit that resonated with me was his reference to bass feeling and speed from his model B. I can certainly relate to bass speed but have found dipole midbass (above 70Hz) to be a bit thin and I yearn for more feeling. Maybe I need 3 or 6x 12" per side vs one 18"...or maybe it's the Faital driver I'm using.
I have 8 x AE Dipole 12 that I’m going to be using in a OB / horn speaker I am putting together, so 4 per side to cover up to c.600hz. I’m currently thinking about the best way implement them.
So I am thinking I might try to clone the Model B woofer configuration albeit with 4 woofers rather than 6, but I’m puzzled by a couple of things; in the video PL says that they load the woofers and squeeze the sound out. This sounds like the same principle as a SLOB or Ripole, but they clearly aren’t using either of those configurations, so how are they doing it? Second, why are the woofers offset and what is happening in the space to the side of them? Perhaps the answer to that also answers my first question ??
if you have a 24 inch wide cabinet; you have plenty of wooferin’ mounting space.
With DIY- you can Go nuts!
From the fellow who had 24 Dayton RSS390 HF subwoofers:
With DIY- you can Go nuts!
From the fellow who had 24 Dayton RSS390 HF subwoofers:
Last edited:
Hi Charlie,
Interesting that you mention the 2268H, I was looking at it when searching for alternatives for the Dipole 18. May I ask for more info and your experiences?
Thanks a lot in advance!
All the best
Mattes
For all the people living in apartments, including myself, it would be great to see your take on a fully dipole system due to much lower bass leakage. @cowanaudio mentioned working on something like this.
For the line arrays in the living room I am considering 9x the ScanSpeak 26W-8534g00 https://www.scan-speak.dk/product/26w-8534g00/ per side. The lower Sd and xmax are compensated by the number of drivers, all still within my budget. More important, I know daudio.nl uses these drivers in the c1 https://daudio.nl/c1/
I am looking forward to the audioXpress article!
For the line arrays in the living room I am considering 9x the ScanSpeak 26W-8534g00 https://www.scan-speak.dk/product/26w-8534g00/ per side. The lower Sd and xmax are compensated by the number of drivers, all still within my budget. More important, I know daudio.nl uses these drivers in the c1 https://daudio.nl/c1/
I am looking forward to the audioXpress article!
Looks like you are not decided to use box subs or not. Answer this first and than choose drivers. To answer this you have to know if you want high SPL at 20hz or just moderate is enough. Room size and distance to front wall.And listening distance to speakers. With open baffle down to 20hz at listening position you will be limited by the above. For that also W frame force cancelation mounting necesary. So many constrains. But if done the impact of low end is huge. Sound is breathing into your chect. From the box sub is the power feel but flat feeling without proper modulation.if compared. Also if dipole at low end you have to boost not only for flat but even more if you design for free field reference.Because psychoacousticaly you judge loudness by all amount not by direct sound. So you get about 4.7dB power less into the room then with box. Actually more diference because box speakers will boost output when against the wall. So in reallity it is very inpractical open baffle very down but if there is posibility the result is impressive if you do correctly. I use down to 20hz but my listening distance is up to 2m, otherwise i would need huge speakers.
So conclusion. Use open baffle just for special case. Use box sub for many reasons 👍
So conclusion. Use open baffle just for special case. Use box sub for many reasons 👍
@Baffless : I definitely plan to use sealed subwoofers.
@billshurv : The LX521 uses the L26RO4Y. It is different than the ROY version, probably better all around. The driver is still in production.
@billshurv : The LX521 uses the L26RO4Y. It is different than the ROY version, probably better all around. The driver is still in production.