Indeed. Levitsky, designer of the LM10n and LT3 tweeter, implemented these in his Consonus Tide speaker and crosses them at 2.2kHz.
The aspect I'm still thinking through is the dynamic range/headroom these drivers would allow. For sure the PTT6.5M08 would be higher, likely "unconstrained" from a practical point of view for my application. Not sure for the LM10n, and I'm lost as to how to assess this, short of pointing to the Tide and it's SPL specs.
You could calculate max SPL from specs: 94 dB/w/m sensitivity. 100W max long term power. Assuming linear relationship between power & SPL, at 100W, SPL should be 113~114 dB. There could be compression as max SPL is approached, but this is likely true of any driver, including the PTT6.5M08. OTOH, with a low limit of just 400Hz (which doesn't require much excursion), there's probably very little compression to max SPL.
I've played around enough with variants of these magnetic planar drivers (Neo 10, 8, 3 & GRS version of the same) to feel dynamic range for mid/high range (say above 300Hz LR4 high pass) is not a significant limitation for home use.
Yeap. Narrow/no baffle is the plan.
How would you think through the best tweeter option, such as Radian LT3, Mundorf AMT 25D, or other, to best match LM10n?
This is great to read!!
Of course I'm not forgetting @cowanaudio comment on post #33 where he expressed a different opinion. At least there is no unanimous opinion that this is not a good idea 🙂
Depends on target Max SPLs.
The 300Hz high pass I mentioned is probably overoptimistic. I've used a Neo 10 as low as 400Hz LR4 and though I had to juggle the crossover & EQ a bit to make it work, it did not strike me as distorted or strained in the mids upwards of 100 dB SPL.
A friend used 2 Neo10s per side for 300~400Hz on up; another used 6 or 8 Neo10 with at least as many Neo3 for treble in line array OB.
My next OB project will probably use a SB Audience Nero-12MWN700D 100~700 Hz, Neo10 or Neo8 to 2.5~3.5 kHz, then Neo 3 or Aurum Cantus 2560 AMT up top. Below 100Hz, a SB Audience Nero-18SW1900D in H frame or dual SB34NRX75-6 in W frame. I have come to really like the rich full delivery with a large nude or near-no-baffle driver in the low mid band.
You ought to find out where the Consonus crosses from its 9.5" to the LM10.
That speaker looks close to what I've been plotting. 😎
Radians seem to have closed back, as well as Mundorf and Dayton and HiVi and some GRS. It is possible to remove the backplate for dipole use.
GRS planars are remakes of Bohlender & Graebner that I use, so 8/10" with 3½" 400Hz xo might be too low, most likely your cone woofer can easily to to 600-800Hz
https://www.parts-express.com/speak...-midranges-tweeters/planar-ribbon-transducers
Going back to the original question:
Some I’ve used in OB speakers and like:
I personally like metal-cone drivers in OB midrange. Especially if crossing to an AMT or hard-dome tweeter, the metal cones can “keep up” and have a more seamless transition.
Some I’ve used in OB speakers and like:
- Eton Arcosia 7”. Some say it doesn’t measure well, and it is a bit overpriced, but the Arcosia drivers really do sound great: smooth and pleasing but also very detailed.
- EAD Sweden E100HD Mkii: Extremely revealing, more-so than the Arcosia even. Sound is thinner/airier with more metallic harshness in the breakup compared to the others. (Their website seems to be down at the moment)
- Markaudio MOAP 10.2: excellent mids & vocals. I slightly prefer the Arcosia & EAD but this performs pretty close to them.
This is only really possible with a very small number of moving coil drivers. They must have a small diameter, e.g. 5.25" and be used nude. The only other alternative is a large planar midrange, but the two models that are available (GRS PT5010 and Radian LT10) start to have vertical directivity problems above 1.5kHz. In the horizontal plane directivity broadens around 1.5k, then returns to a more of a dipole but then starts to worsen around 4kHz. This is all when it is used nude. Putting it in a baffle even a small one makes these effects worse. Also, using the LT10 below about 600Hz is not advised because this is where distortion is starting to rise and dipole losses are getting larger.
I have owned and measured both the LT10 and GRS PT5010 as well as some AliExpress clones.
For these reasons, and because the planars are more delicate than a moving coil driver, I don't really see these planars as a "better" alternative to a good moving coil driver.
Wow that's low... Consonus Tide uses many basic designs that are similar to my AINOgradients. Looks like Consonus is in USA. Price is 80.000$/pair including dsp+amps
https://audioxpress.com/news/new-audio-brand-consonus-debuts-tide-speaker-system-at-axpona-2024
They say they are pending a patent for crossing monopole sub to dipole woofer - but I see no point/success in that. That has been done for ages by Quad/Gradient, Martin Logan etc. It is based on simple acoustic principles.
https://audioxpress.com/news/new-audio-brand-consonus-debuts-tide-speaker-system-at-axpona-2024
They say they are pending a patent for crossing monopole sub to dipole woofer - but I see no point/success in that. That has been done for ages by Quad/Gradient, Martin Logan etc. It is based on simple acoustic principles.
Last edited:
Thanks for including these links! I have a fair amount to read and digest.Radians seem to have closed back, as well as Mundorf and Dayton and HiVi and some GRS. It is possible to remove the backplate for dipole use.
GRS planars are remakes of Bohlender & Graebner that I use, so 8/10" with 3½" 400Hz xo might be too low, most likely your cone woofer can easily to to 600-800Hz
https://www.parts-express.com/speak...-midranges-tweeters/planar-ribbon-transducers
My 6.5" woofers can go to 800Hz, however the cavity in the SLOB makes their response drop/cut at 470Hz. I might be able to extend it some more. I will try it with my current setup.
Here's how the SLOB measures on post #1035: https://www.diyaudio.com/community/threads/the-xsd-speaker.385717/page-52
Great input! And you have heard/measured the LM10n and PTT6.5M08 as well as many others.
Gets to the core of the issue I'm trying to figure out.
Your comment "using the LT10 below about 600Hz is not advised because this is where distortion is starting to rise and dipole losses are getting larger." made me recheck Dickason measurements, copied below.
Help me understand: on the left, H2 and H3 exhibit significant rise until 300Hz and lower. But the CSD does show something is going on between 300 and 1000Hz. I'm confused as I would expect distortion to increase aligned with CSD, or CSD to behave below 1kHz as it does above 1kHz, but not one going one way and the other in the opposite way?
@LewinskiH01 Oops, I made a mistake! What I bought, tested/measured was the Radian LM8k not the LM10n! I only realized this after going back and looking for my measurements to post here. So the LM10n might be just fine when used lower than 600Hz. But for the PT5010, which I believe is similar to the LM10n, the FR is falling off and the distortion rising already around 500Hz (for a nude application). After you boost up the SPL to match the higher frequencies, distortion will also rise and this is why I (in my memory) felt that 600Hz was a good "not lower than this" limit the the usable/practical frequency response.
But overall this doesn't really change my feeling about planar midranges and OB/dipole applications. Moving coil drivers can go lower, take more power input, often have lower distortion, and IMHO are more reliable. But it really depends on exactly what you want/need to get out of the midrange and where that band stops and ends in your own design.
But overall this doesn't really change my feeling about planar midranges and OB/dipole applications. Moving coil drivers can go lower, take more power input, often have lower distortion, and IMHO are more reliable. But it really depends on exactly what you want/need to get out of the midrange and where that band stops and ends in your own design.
Indeed. When I saw the Tide design I very much thought about your AINOgradients.
From your writeup/pdf it looks like you cross your midrange at around 450 and 1700Hz? From fig. 3.14 on page 7.
Do you happen to have measurements for your midrange within that passband that you can share? Most interested in distortion, given the discussion above.
Tide mid/bass: This video explains some more about the Tide design:
While my 4 sealed subs are in the 4 corners of my room (instead of by the speakers), I want to experiment with this approach and hear how cardioid sounds in my case. Do you think the 4-carner subs would defeat cardioid implementation?
Charlie,
Making direct comparisons between the Parts Express planar drivers and the Radians is as little unfair. They may be similar designs, but everything about the construction and quality of the Radians is superior. PE are poor copies of the BG designs. Radian are fully optimized designs, improving upon the original BG units.
Stronger magnets, stronger chassis, more precise stretching of the diaphragms. Better consistency from unit to unit. In my own testing between original BG NEO 10 and the Parts Express, they are not even close. Much higher distortion from the Parts Express. Same goes for the 8" versions.
OK, but I did buy the RADIAN LM8k and measured it, just not the bigger one. The RADIAN design is definitely beefier/heavier but it hasn't really changed my mind about planars.
Here is my data on a nude LM8k, taken at a moderate SPL level... it's already exceeding 1 percent distortion around 800 Hz and getting worse below that. I do not expect the LM10n to be any better below 500-600Hz but I have not measured one myself so I can't say with 100% certainty. Yes the RADIAN units are better built, but the performance seems to be on par with the original B&G designs and not breaking any new ground IMHO.
Here is my data on a nude LM8k, taken at a moderate SPL level... it's already exceeding 1 percent distortion around 800 Hz and getting worse below that. I do not expect the LM10n to be any better below 500-600Hz but I have not measured one myself so I can't say with 100% certainty. Yes the RADIAN units are better built, but the performance seems to be on par with the original B&G designs and not breaking any new ground IMHO.
Attachments
Last edited:
That seems counter to what's known about these magnetic planars. Maybe it's the trapezoid baffle around the LM10n that helps it go so low? Or is it some secret sauce in the crossover?
OTOH, this is a speaker by the original designer of the LM10n. What does he know that we don't?
Indeed, 220 Hz is surprisingly low. The trapezoid baffle will certainly help a bit. The Consonus product brochure claims they can do 114 dB short term.
I could see the 220Hz Xover point as workable given the baffle width IF they use a low order crossover, for which the attenuation starts out at a higher frequency and could be -6 dB down at the crossover point... e.g. LR2. See:
https://en.wikipedia.org/wiki/Linkwitz–Riley_filter#/media/File:Linkwitz_vs_Butterworth.svg
https://en.wikipedia.org/wiki/Linkwitz–Riley_filter#/media/File:Linkwitz_vs_Butterworth.svg