Offer to fund driver purchase for FAST design with ASR rigor

[Ilkless]

Those who follow my posts would know I think there's some untapped potential for a speaker with woofer-assisted wideband driver (a.k.a FAST) as a design format. That's not helped by the rampant voodoo in the space. That's a shame as I think there's a niche for a well-engineered design with intelligent compromises if a modern widebander is used. Think Directiva but really old-school.

Thinking a 3-4" widebander mated to an 8+10" woofer with a 400-600Hz crossover.

Potential advantages:

- Better vertical lobing due to low XO point for the CtC

- Directivity control due to wide drivers + baffle without resorting to open baffle/slots ala Geithain or DD. Perhaps there is room to optimise CtC, driver diameters, baffle width and height for the smoothest directivity within this design format. The closest a wide baffle speaker, panels aside, came to mainstream acceptance were the Sonus Faber Elipsa and Devore Orangutan. And I suppose, the Linkwitz LXMini for a woofer-assisted widebander, or the XRK971 transient-perfect FAST for DIYA denizens.

- Less compression and more SPL than smaller monitors

- Not as loud as compression drivers but smoother

- More affordable, simpler construction, less to screw up in design and construction than typical large woofer direct radiating speakers with 3+ ways

- Less directivity mismatch than 8-10" 2-ways with a dome tweeter (looking at the Devore... the mismatch was painfully obvious to me)

Compromises

- Slightly less smoothness and HF extension than a dome tweeter, potentially rapidly narrowing directivity (extent of narrowing depending on the exact driver used)

- Potentially IMD in midrange and HD in HF

- Size/spouse approval factor

Broad design goals

- 3-4" widebander, 8-10" woofer

- 400-600Hz XO

- No preference for what box alignment, be it TL, reflex, aperiodic, sealed

- Explore the possibility of slightly waveguiding the widebander

- 35-16kHz +/- 1.5dB

- <1% THD 100Hz-16kHz at 100dBSPL

- Low baffle step frequency and smoothly increasing directivity on spinorama, established first with sims and then with anechoic/well-done gated/free-field/NFS measurements

- Broadly accessible drivers

The obvious candidate widebands I know of are the:


- Scanspeak 10F/8414G10

Scanspeak 10F/8414G10 Discovery 4 inch Fullrange

Scanspeak 10F/8414G10 Discovery 4 inch Fullrange

www.erinsaudiocorner.com

- Vifa TC9/TG9

- SEAS FU10RB

- SB Acoustics SB65 if using a woofer on the smaller side

- Mark Audio widebanders

I will pay for wideband + woofer drivers plus shipping, up to around $250, for a single prototype unit. I have nothing in it, just to spark discussion and collaboration in the community, and to bring much-needed ASR rigor to this design format and see what the results can be, as a member from a country where DIY speakers are prohibitive.

 

[mcdn]

Perhaps the modern widebander is the tectonic BMR, e.g. the https://www.tectonicaudiolabs.com/product/tebm46c20n-4b/ , which is used in the fairly wide baffled Philharmonic BMR to great effect: https://philharmonicaudio.com/Measurements-BMR.html

Dennis uses a ribbon tweeter to fill in the top and a smallish woofer, but a larger woofer version without the tweeter would probably get a long way the classic sound and look.

 

[Ilkless]

Perhaps the modern widebander is the tectonic BMR, e.g. the https://www.tectonicaudiolabs.com/product/tebm46c20n-4b/ , which is used in the fairly wide baffled Philharmonic BMR to great effect: https://philharmonicaudio.com/Measurements-BMR.html

Dennis uses a ribbon tweeter to fill in the top and a smallish woofer, but a larger woofer version without the tweeter would probably get a long way the classic sound and look.

I recall reading @Dennis Murphy saying he did not like the top end of the Tectonic without the ribbon, and my other concern is with power handling. Though I'm looking forward to what he will do with the new, higher end Tectonic BMR.

There are other wideband drivers from pro audio crossover brands, like Sica, Faital and PRV. There might also be a suitable candidate from there but I'm not as familiar with the exact offerings from those brands.

 

[mcdn]

I recall reading @Dennis Murphy saying he did not like the top end of the Tectonic without the ribbon, and my other concern is with power handling. Though I'm looking forward to what he will do with the new, higher end Tectonic BMR.

There are other wideband drivers from pro audio crossover brands, like Sica, Faital and PRV. There might also be a suitable candidate from there but I'm not as familiar with the exact offerings from those brands.

All true, but every wideband option is going to be compromised somehow. Let's say we're talking about a crossover at 500Hz. I don't think there are any drivers that can do 500-20k without any issues, so you have to choose your poison. Or add a tweeter, but then it doesn't meet your design goal.

 

[abdo123]

All true, but every wideband option is going to be compromised somehow. Let's say we're talking about a crossover at 500Hz. I don't think there are any drivers that can do 500-20k without any issues, so you have to choose your poison. Or add a tweeter, but then it doesn't meet your design goal.

The scanspeak driver recommended at the very beginning seems very capable.

 

[mcdn]

The scanspeak driver recommended at the very beginning seems very capable.

It inevitably beams like heck above 5k

 

[abdo123]

It inevitably beams like heck above 5k

Not really, i’ve seen worse from regular tweeters.

 

[thewas]

It inevitably beams more than a typical decent tweeter in the last 2 two octaves, also for a tweeter a waveguide can be easier used to unify its directivity more (make it narrower at lower and wider at higher frequencies).

For example comparing the 60° at 10 kHz, the 4" Scanspeak has dropped already 15 dB

(source: https://www.erinsaudiocorner.com/driveunits/scanspeak_discovery_10f8414g10_fullrange/)

while a 1" dome only around 12 dB

(source: https://www.erinsaudiocorner.com/driveunits/peerless_dx25tg59-04_tweeter/)

or 8 dB

(source: https://www.erinsaudiocorner.com/driveunits/seas_27afncd_tweeter/)

which could be further reduced with a WG or a smaller tweeter.

In my experience the abrupt kink in directivity in the presence region and lack of sound power in the last octaves is quite audible, so despite liking the simplicity of such FAST designs I wouldn't make one anymore, the reason why for example also good loudspeaker engineers still use a tweeter above a small and relatively wide radiating BMR driver.

 

[Scgorg]

Don't really have anything to add here, other than agreeing that a sensible design should be possible within the constraints presented. A good widebander candidate I recently came across is the Kartesian Wib120_vHE (https://www.kartesian-acoustic.com/wib120-vhe). It has very nice large-signal parameters for the application (very linear and symmetrical Le(x), and very linear BL(x) within +-2mm) which should imply low intermodulation distortion. It also extends quite cleanly to 20kHz with an off-axis response that looks passable until 10kHz at least.

The most obvious negative of the driver is the cost, which is about 150 euros per driver from TLHP. It's also on the larger side, so it will naturally be more directional than some other alternatives.

 

[Ilkless]

For example comparing the 60° at 10 kHz, the 4" Scanspeak has dropped already 15 dB

(source: https://www.erinsaudiocorner.com/driveunits/scanspeak_discovery_10f8414g10_fullrange/)

while a 1" dome only around 12 dB

(source: https://www.erinsaudiocorner.com/driveunits/peerless_dx25tg59-04_tweeter/)

or 8 dB

(source: https://www.erinsaudiocorner.com/driveunits/seas_27afncd_tweeter/)

which could be further reduced with a WG or a smaller tweeter.

In my experience the abrupt kink in directivity in the presence region and lack of sound power in the last octaves is quite audible, so despite liking the simplicity of such FAST designs I wouldn't make one anymore, the reason why for example also good loudspeaker engineers still use a tweeter above a small and relatively wide radiating BMR driver.

The point is to see how that directivity kink can be avoided, by allowing a simple 2-way large woofer design without as large a mismatch as those that use a dome tweeter (e.g. the Devore O/96). And the directivity control in the midrange that comes with a larger woofer + baffle. I've owned and heard several speakers with 4" wideband and I felt the narrow directivity about 10kHz was acceptable so long as the roll off is smooth.

 

[thewas]

The point is to see how that directivity kink can be avoided, by allowing a simple 2-way large woofer design without as large a mismatch as those that use a dome tweeter (e.g. the Devore O/96). I've owned and heard several speakers with 4" wideband and I felt the narrow directivity about 10kHz was acceptable so long as the roll off is smooth.

I am talking about the directivity kink from the wideband driver itself, I have heard and even used in the past myself several such decent 3-4" drivers and was always missing the sound power in the last octaves. I guess it is the same reason why companies like Nubert and loudspeaker designers like @Dennis Murphy use small and wide radiating BMRs also only as mid drivers. But if you were happy in the past with their sound, why not?

 

[Ilkless]

I am talking about the directivity kink from the wideband driver itself, I have heard and even used in the past myself several such decent 3-4" drivers and was always missing the sound power in the last octaves. I guess it is the same reason why companies like Nubert and loudspeaker designers like @Dennis Murphy use small and wide radiating BMRs also only as mid drivers. But if you were happy in the past with their sound, why not?

Ah, yes, there are directivity kinks from breakup in drivers in the presence region like the Jordan JX92 for sure, but the Scanspeak for example seems to have really well controlled (just slightly narrower) directivity until it narrows significantly past 10kHz.

 

[thewas]

Ah, yes, there are directivity kinks from breakup in drivers in the presence region like the Jordan JX92 for sure, but the Scanspeak for example seems to have really well controlled (just slightly narrower) directivity until it narrows significantly past 10kHz.

The problem of the Scanspeak is that its directivity is very wide till 3 kHz and then rises, in my experience a directivity kink at such a frequency is quite audible, good typical multi-way loudspeakers have it quite lower.

 

[Ilkless]

The problem of the Scanspeak is that its directivity is very wide till 3 kHz and then rises, in my experience a directivity kink at such a frequency is quite audible, good typical multi-way loudspeakers have it quite lower.

Quite lower as in directivity starts to narrow from a lower frequency?

 

[Ilkless]

Linkwitz LXmini Speaker Review

Linkwitz LXmini Review

www.erinsaudiocorner.com

LXMini spinorama and DI looks decent until 16kHz, it would probably perform better with a flat baffle.

 

[alex-z]

Directivity control via a large woofer and large baffle is not terribly practical. Speaking from experience using a 27x49" baffle for the front L/R in my main setup.

Such large baffles come with a serious downside and that is resonance. To make mine tolerable, I have 1/2" plywood sandwiched with 3/4" MDF using green glue. The legs are 2x4 lumber triangles, and 80lbs of concrete blocks holding them down.

If you are willing to go through such effort, using a coaxial plus 3 way crossover just makes more sense than a 2 way FAST. It would be a shame to have nice low frequency directivity and then have it be uncontrolled at the high end.

 

[mcdn]

Linkwitz LXmini Speaker Review

Linkwitz LXmini Review

www.erinsaudiocorner.com

LXMini spinorama and DI looks decent until 16kHz, it would probably perform better with a flat baffle.

I run LXMinis in our family room so I listen to them every day. The DI is very smooth, but they require a lot of DSP to get the FR right. If you are willing to go active with DSP then a lot more options open up.

 

[thewas]

Quite lower as in directivity starts to narrow from a lower frequency?

Yes, also the change of the slope is important, ideally smooth and as far away as possible from the mid band.

 

[KSTR]

Tang Band W4-1320 is a very nice wide-band driver, measurements here: http://rutcho.com/speaker_drivers/tang_band_w4_1320_sb/tang_band_w4_1320_sb.html. Due to the phase-plug the beaming is quite moderate. The inherent damping in the bamboo-fibre cone also helps a lot for smooth HR response and fast decay, compared to its metal-cone brother, the W4-1337.

Personally, I find a widerbander's best use is small, bandwidth-restricted near-field (1m, max) monitor, preferably active design (current drive at MF/HF). Spoken word and singing voice naturalism and clarity is often breath-taking, same for small acoustic sets. Stereo imaging is stunningly impressive when mounted flush on "infinite" baffle, no diffractions and reflections.
But they can't go loud, even with an assisting woofer, unless helped by a really huge waveguide/horn, also flattening directivity. Goodbye to simplicity at this point...

 

[Martin]

The Von Schweikert Unifield 3 uses a similar concept. I attended the 2009 RMAF with Albert Von Schweikert who was featuring the Unifield 3 at the show. He called it an augmented full range speaker. The full range speaker he employed was a 5” Fostex driver. It was crossed over to a 7” woofer at 100Hz and a ribbon tweeter at 8kHz. It is a wonderful sounding speaker which has all the benefits of a full range speaker with bass and treble extension most lack. I often played with the idea of building something similar with a larger woofer (maybe 12”) to give more bass extension and impact.

Martin