Do you find that the rear firing tweeter makes the listening room “sweet spot” a lot bigger, given the dispersion pattern of the 15” FR driver?
Perry, in the other thread you said of the Bitches Brew: “2 1/2 years later I'd be hard pressed to significantly improve them.”
I’ve just watched an interview with Peter Lyngdorf in which he said the new 10” Purifi woofer has far and away the lowest IMD of any driver ever made, irrespective of size. He doesn’t strike me as someone given to BS and everything we have seen from Purifi so far makes this a credible claim. So maybe an OB based on this new woofer might be something to explore ? I know Charlie Laub also has it in his sights.
I’ve just watched an interview with Peter Lyngdorf in which he said the new 10” Purifi woofer has far and away the lowest IMD of any driver ever made, irrespective of size. He doesn’t strike me as someone given to BS and everything we have seen from Purifi so far makes this a credible claim. So maybe an OB based on this new woofer might be something to explore ? I know Charlie Laub also has it in his sights.
I agree. Haven't tried them but I don't doubt it's true. Purifi 10 would be a great midrange in an application like this. You could use it as a bass but I think you'd need multiples, as generally I wouldn't do a bass dipole with a driver smaller than 12".
Of course I did indeed have my eye on the PTT10s but my application requires eight of them. They cost about $675 each - do the math and keep in mind this is without shipping! I really do not think that several thousand dollars on these drivers is a good idea at this point, even if they are "the best ever" 10 inch.
I built the speakers already using some nice Scan-Speak 10" woofers that have almost identical TS parameters but Xmax is only 6mm. The PTT10s have about 3x the Xmax but this is only going to get me around 3-4dB more output and perhaps a bit lower distortion. The concept behind the speaker has been proven to work well, so there is no urgency for the large outlay at this point.
I built the speakers already using some nice Scan-Speak 10" woofers that have almost identical TS parameters but Xmax is only 6mm. The PTT10s have about 3x the Xmax but this is only going to get me around 3-4dB more output and perhaps a bit lower distortion. The concept behind the speaker has been proven to work well, so there is no urgency for the large outlay at this point.
I use the PTT6.5M in the speaker I mention above and it is really great as a midrange. Not sure I would use a 10" unless you can cross over to another driver below 1k Hz. The response pattern will start to narrow around 900Hz or so with any 10" due to the cone diameter.
I believe the Purifi 10 or its 8" cousin would radiation pattern match very nicely with the SB Acoustics 29" Beryllium tweeter with waveguide, and in a minimally baffled [for the midwoofer] dipole yield constant directivity and a great looking polar heat map from 100Hz-20Khz.
@perrymarshall, thank you for these great designs! Question on the Radian: can you buy the driver with aluminum dome and upgrade to beryllium later? Also, with the price of the beryllium version, does is still makes sense to go with Live Edge Dipoles vs Bitches Brew?
YES you can absolutely start with aluminum and upgrade later. (It was even cheaper for me to do it that via upgrade when I bought them. Your mileage may vary.) The Beryllium Live Edge Dipole is less expensive overall than the Bitches Brew design.
It is insanely good. I used it in these speakers and I describe it in some detail here:
https://psma-website-assets.s3.amazonaws.com/live_edge_beryllium_dipoles.pdf
The TN29BNWG Beryllium waveguide tweeter crosses at 1500Hz to a Lavoce 15" woofer and the radiation patterns match very nicely. I think it would match the Purifi 10 well.
So, a couple questions:
I haven't seen it mentioned in the articles or couple threads I've seen on these designs, but what are the requirements as far as placement relative to walls and any other important placement parameters? I guess, what is minimum required to get good sound and dipole performance, and then what is optimal placement?
I did see that it should not be parallel to a wall, which is easy in most any room and generally required with box speakers as well for reasonable imaging.
Second, what does vertical directivity look like?
I haven't seen it mentioned in the articles or couple threads I've seen on these designs, but what are the requirements as far as placement relative to walls and any other important placement parameters? I guess, what is minimum required to get good sound and dipole performance, and then what is optimal placement?
I did see that it should not be parallel to a wall, which is easy in most any room and generally required with box speakers as well for reasonable imaging.
Second, what does vertical directivity look like?
1-1.5M / 3-5 feet from walls sounds great. I wouldn't put them much closer. You can more easily get away with being closer to side walls. Because of the dipole null at 90 degrees off axis, there isn't much sidewall reflection.
Above 500Hz Vertical directivity is 100% identical to Horizontal directivity because the drivers are coaxial.
I don't have a good way of measuring Vertical directivity below 500Hz, but informally from taking all kinds of measurements at all kinds of angles, neither system has significant lobing at its lower crossover frequency (100Hz Bitches Brew, 200Hz Live Edge Dipoles). Very well behaved.
Above 500Hz Vertical directivity is 100% identical to Horizontal directivity because the drivers are coaxial.
I don't have a good way of measuring Vertical directivity below 500Hz, but informally from taking all kinds of measurements at all kinds of angles, neither system has significant lobing at its lower crossover frequency (100Hz Bitches Brew, 200Hz Live Edge Dipoles). Very well behaved.
Thanks! Really nice looking speakers and interesting electo-acoustic design. I'll keep these in mind for when the kids get a little older 😀
I have heard some DIY box speakers that use the SBA Be tweeter (without waveguide) surrounded by an array of 14 x Mark Audio CHN 50 full range drivers. The latter only handle the mid range frequencies and hand off to a pair of woofers. Needless to say the upper frequencies sound fabulous but even more impressive is the mid range which is ESL-like in its purity and transparency. The CHN 50 has a QTS of 0.55 and it has made me think about a 3 way OB using the same tweeter and mid range arrangement and adding a suitable large woofer.
Obviously radiation pattern and room ambience are a big reason why we here like open baffles, but a less obvious reason is that the inherent logic and necessities of a Dipole Design lead you to high efficiency and large amounts of surface area.
For a long time, I was obsessed with compact, low efficiency designs, and I explored that territory thoroughly. Then OB with Professional drivers with 97 dB sensitivity introduced a whole new dimension of dynamics to the music.
In the 80s, I was turned off by horn designs because most of them had ragged frequency response etc. but horns have vastly improved because of computer modeling, and when you can fix all of their inherent problems with DSP, you’re in a whole new era.
The system you’re describing invokes all of that so long as the full range drivers are not burdened with low frequencies. Waveguides effectively triple the surface area of a tweeter and give it those same advantages.
I have also used the SB acoustics standard 29mm beryllium tweeter and it is great just like you say.
For a long time, I was obsessed with compact, low efficiency designs, and I explored that territory thoroughly. Then OB with Professional drivers with 97 dB sensitivity introduced a whole new dimension of dynamics to the music.
In the 80s, I was turned off by horn designs because most of them had ragged frequency response etc. but horns have vastly improved because of computer modeling, and when you can fix all of their inherent problems with DSP, you’re in a whole new era.
The system you’re describing invokes all of that so long as the full range drivers are not burdened with low frequencies. Waveguides effectively triple the surface area of a tweeter and give it those same advantages.
I have also used the SB acoustics standard 29mm beryllium tweeter and it is great just like you say.
“ If you can see the cone moving, it’s distorting . . .” The corollary is that if the cone isn’t moving then there’s no distortion and with 14 drivers per speaker handling just the mid range, their cones are barely moving. I’m convinced that is why the mid range on the speaker I mentioned is so outstanding.
OK I admit that's a bit of an exaggeration. Horns are hard to tame, domes are much more well behaved.
THAT SAID... I am speaking from experience and these designs here are evidence. I'm able to get the exact radiation pattern I want; the blend with other drivers that I want; the way the system is designed is
#1 priority is radiation pattern
#2 priority is power and dynamic range considerations
#3 priority is time and phase
#4 priority is frequency response
Traditionally all system priorities are done around #4 and if you use passive xovers you are forced to make #4 top priority. All of the other departments suffer.
But with DSP that requirement is decoupled. Now #1 and NOT #4 determines the physical configuration and driver choice.
These designs would be difficult if not impossible to pull off w/o DSP. But in these designs, we address 1 2 and 3 through physical configuration and we fix peaks and dips in horn / coaxial response (a coaxial has even more problems than a traditional horn) with DSP. And we fix dipole loss with DSP. In the Live Edge Dipoles I'm using 20dB of bass EQ.
Dome tweeters have lots of compression and lots of distortion and their radiation patterns are problematic. It takes skill to coax the best performance out of a horn and when you do it right it works wonderfully well. I'll match these horn designs against ANY dome tweeter speaker at low SPL levels, and blow them away at high output levels.
A crossover is a powerful thing in some areas, but there are of course limits to what it can do.
I'm not sure what you're implying. You can't change the radiation pattern of a horn or a driver using electronic means of any sort.