Meadowlark Kite Build Thread - Prime Minister Gets A New Pair of Speakers

It doesn't get much better than this, Satori drivers from SB Acoustics.

DSCN1144.jpg



Note the tweeter's surround. It's unusually wide and made of a very light weight fabric rather than rubber. The result is a more linear and freer suspension for better dynamic linearity. That combined with the Beryllium dome result in a very low moving mass. Beryllium is light stuff, element #4. Unlike other metal domes you'll see an absence of ringing at the top of the range. Quite a low Fs considering the mass is so low; that's because of the extra tricky dual back cavities.


TW29.jpg

Notice that the faceplate is two concentric rings that are elastically decoupled.

Tweeter artifacts manifest in the vocal range, so they're usually attributed to something other than the tweeter. This one's as clean as a whistle and that really shows up on female vocal. For that, there's a hard to achieve character we call "Dew on the Bosom", this one gets us there.


DSCN1145.jpgDSCN1146.jpgDSCN1147.jpg

This hansom guy is large for his category with a nice, big effective area and exceptionally long stroke. Stoke times area equals displacement, the most important measure of how much work a woofer can do as frequency descends - a factor we love to exploit with DSP.

SB's gone all-out on suspension symmetry with a flat spider facing wide open air loads on both sides, they've even brought the coil braids out at 180 degrees to negate any possible coil cock. Notice that the motor is elastically decoupled from the basket. That's a first. Plus every coil cooling trick in the book.

Nicely vented pole piece with a flare at both ends. Neodymium motor that is extremely well focused on a tallish gap, terrific flux density. The former looks a bit long, but that's to accommodate the extended copper shorting sleeves on the pole piece. So this motor packs a punch with very low distortion.

When you feel the stroke you'll find the compliance especially soft and linear. The cone is a house specialty made of pressed papyrus - which seems like a weird idea until you hear it. Light, stiff and dead; SB's Scan Speak lineage is showing here. A little cone cry is evident on the bench but it's way, way, way above range and out of the way.

All of these things combined give us a driver with noticeably superior transient response, naturalness in the mids and manliness in the bass. Just gorgeous, this woofer is a powerful "shot across the bow" for every other woofer maker. You should hear it in tandem with its 9.5" big brother.

Any DIYers among us? You have to try this driver! I'd pay twice as much for it, maybe more. Plus it's breezy-easy to use.
 
It doesn't get much better than this, Satori drivers from SB Acoustics.

View attachment 12455



Note the tweeter's surround. It's unusually wide and made of a very light weight fabric rather than rubber. The result is a more linear and freer suspension for better dynamic linearity. That combined with the Beryllium dome result in a very low moving mass. Beryllium is light stuff, element #4. Unlike other metal domes you'll see an absence of ringing at the top of the range. Quite a low Fs considering the mass is so low; that's because of the extra tricky dual back cavities.


View attachment 12459

Notice that the faceplate is two concentric rings that are elastically decoupled.

Tweeter artifacts manifest in the vocal range, so they're usually attributed to something other than the tweeter. This one's as clean as a whistle and that really shows up on female vocal. For that, there's a hard to achieve character we call "Dew on the Bosom", this one gets us there.


View attachment 12456View attachment 12457View attachment 12458

This hansom guy is large for his category with a nice, big effective area and exceptionally long stroke. Stoke times area equals displacement, the most important measure of how much work a woofer can do as frequency descends - a factor we love to exploit with DSP.

SB's gone all-out on suspension symmetry with a flat spider facing wide open air loads on both sides, they've even brought the coil braids out at 180 degrees to negate any possible coil cock. Notice that the motor is elastically decoupled from the basket. That's a first. Plus every coil cooling trick in the book.

Nicely vented pole piece with a flare at both ends. Neodymium motor that is extremely well focused on a tallish gap, terrific flux density. The former looks a bit long, but that's to accommodate the extended copper shorting sleeves on the pole piece. So this motor packs a punch with very low distortion.

When you feel the stroke you'll find the compliance especially soft and linear. The cone is a house specialty made of pressed papyrus - which seems like a weird idea until you hear it. Light, stiff and dead; SB's Scan Speak lineage is showing here. A little cone cry is evident on the bench but it's way, way, way above range and out of the way.

All of these things combined give us a driver with noticeably superior transient response, naturalness in the mids and manliness in the bass. Just gorgeous, this woofer is a powerful "shot across the bow" for every other woofer maker. You should hear it in tandem with its 9.5" big brother.

Any DIYers among us? You have to try this driver! I'd pay twice as much for it, maybe more. Plus it's breezy-easy to use.

Man, that's driver porn. Easily the most impressive drivers I've ever seen. Can't wait to hear them.

Btw, are those ones mine? Not that it really matters, but the 4 year old in me needs to know. :)
 
Man, that's driver porn. Easily the most impressive drivers I've ever seen. Can't wait to hear them.

Btw, are those ones mine? Not that it really matters, but the 4 year old in me needs to know. :)
Yep. Just got 'em. Clean machines! Nothing matters more than the transducers.
 
The Scan Speak 18W/8545 in Shearwater is really a classic design that is, remarkably, still in production with only slight improvements. In the 90s a lot of us were using it. Dave Wilson, Bobby Palkovic and Paul Hales come to mind.

You'll notice that the emissive area is relatively large for the frame diameter, always nice. And there's plenty of motor; this predates neodymium. The spokes are far from aerodynamic, quite the opposite, really. And the spider is typical for that period: asymmetrical and non venting. Pretty conventional all the way round, but heavy duty and built with great care. Plenty of flux density resulting in good motor force on a light moving system.

What's most interesting is the cone design. It's follows along on an idea originated, I think, by Bozak: the diaphragm is thickest where it joins the former then thins progressively toward the edges. The idea was to defeat break up with flexure. That concept later became anathema as stiff cones won the day. But it did work, albeit giving some ground at the bottom of the band. And there was one giant plus that we took advantage of in Shearwater - because the cone flexed more with ascending freq the emissive area effectively became smaller with ascending freq, greatly lessening beaming, so it could be used to higher freqs without ruining the room response; an ideal situation for a first order two way.

20190415_085939.jpg

20190415_090058.jpg
 
The Scan Speak 18W/8545 in Shearwater is really a classic design that is, remarkably, still in production with only slight improvements. In the 90s a lot of us were using it. Dave Wilson, Bobby Palkovic and Paul Hales come to mind.

You'll notice that the emissive area is relatively large for the frame diameter, always nice. And there's plenty of motor; this predates neodymium. The spokes are far from aerodynamic, quite the opposite, really. And the spider is typical for that period: asymmetrical and non venting. Pretty conventional all the way round, but heavy duty and built with great care. Plenty of flux density resulting in good motor force on a light moving system.

What's most interesting is the cone design. It's follows along on an idea originated, I think, by Bozak: the diaphragm is thickest where it joins the former then thins progressively toward the edges. The idea was to defeat break up with flexure. That concept later became anathema as stiff cones won the day. But it did work, albeit giving some ground at the bottom of the band. And there was one giant plus that we took advantage of in Shearwater - because the cone flexed more with ascending freq the emissive area effectively became smaller with ascending freq, greatly lessening beaming, so it could be used to higher freqs without ruining the room response; an ideal situation for a first order two way.

View attachment 12584

View attachment 12585

Interesting stuff!
So what did neodymium get us? Also, are aerodynamics really that important?
 
Interesting stuff!
So what did neodymium get us? Also, are aerodynamics really that important?
My understanding is that neo was developed by the car guys who wanted to shrink the size of starters. Higher flux, smaller size. Cram that mill into that car!

For drivers, higher flux density translates into higher sensitivity or the ability to lengthen the coil (for more stroke) or some combination of the two. So, yeah, very desirable. The low power guys want sensitivity all by itself. I'm in the "power is cheap so use lots" camp. But, either way, better sensitivity means less current passing thru the coil, so less heat, so less thermal compression, better dynamics.

Drivers like the Satori pretty much split the difference - so we get both cooler coils and more stroke. Yay!

You want the diaphragm to "see" a symmetrical load both fore and aft. Making the spokes as streamlined as possible is part of that. Aside from drag, there's a problem with reflections. Here's a what Scan Speak has done with the Illuminator series to try to make the motor more aerodynamic.

20190415_153418.jpg
 
While It doesn't entirely surprise me, I'm still amazed at the science behind speaker design. More factors involved than I'd really thought about. Thanks, Pat, for the walkthrough. And, I'm guessing that you're just scratching the surface too. Your passion for audio is so much fun. It's what captivated me, separated you from all the other brands in my mind, and made me such a fan. I'm still saving. I just hope that I can save enough to catch you before you retire.

Do you still engrave your meadowlark logo someplace into your enclosure? I've always loved that. Such an awesome thread.
 
My understanding is that neo was developed by the car guys who wanted to shrink the size of starters. Higher flux, smaller size. Cram that mill into that car!

For drivers, higher flux density translates into higher sensitivity or the ability to lengthen the coil (for more stroke) or some combination of the two. So, yeah, very desirable. The low power guys want sensitivity all by itself. I'm in the "power is cheap so use lots" camp. But, either way, better sensitivity means less current passing thru the coil, so less heat, so less thermal compression, better dynamics.

Drivers like the Satori pretty much split the difference - so we get both cooler coils and more stroke. Yay!

You want the diaphragm to "see" a symmetrical load both fore and aft. Making the spokes as streamlined as possible is part of that. Aside from drag, there's a problem with reflections. Here's a what Scan Speak has done with the Illuminator series to try to make the motor more aerodynamic.

View attachment 12600

Hmmmm.
It sounds from your description like high sensitivity is an all around good thing. So why trade off some of it and use all the Watts? Aren't you needlessly complicating the amplifier side? Wouldn't you gain in low level, low power listening with the added efficiency? How about for folks who tastes run more towards the small and quiet end of the musical spectrum? They won't need all those Watts.
 
Hmmmm.
It sounds from your description like high sensitivity is an all around good thing. So why trade off some of it and use all the Watts? Aren't you needlessly complicating the amplifier side? Wouldn't you gain in low level, low power listening with the added efficiency? How about for folks who tastes run more towards the small and quiet end of the musical spectrum? They won't need all those Watts.
Ha! We really don't concern ourselves with "small and quiet". Those needs have been met long ago. Our pursuit is the frontier of realistic reproduction of even the most energetic performances. Of course, you really can't expect to get there with a bookshelf speaker because, among many reasons, you can't transduce the bottom half of the bass. That takes plenty of driver and gobs of juice. For that we simply let you add LF system, as much as you like.

I should point out that having both bandwidth and dynamic headroom in no way lessen the system's ability to play "small and quiet". So those needs are still nicely met. And since, at those levels, the system is barely idling, there's a sense of unrestrained ease.

And, frankly, amp power is cheap so there really is no reason to avoid using it. A 500W N-Core, and absolutely divine hunk o' amp, costs about $100. I can save a few bucks by backing off to the 100 Watter, but why? Then the user would be able to clip his amp during vigorous play; that ain't good. I'd much prefer to have the 'power limit' issue moved up and completely out of play. We use enough power to fry the coils, then prevent that from happening in the processor.

Even though we don't care much about amp power, high sensitivity DOES still mean we use less of it. But the real advantage is cooler coils.
 
Ha! We really don't concern ourselves with "small and quiet". Those needs have been met long ago. Our pursuit is the frontier of realistic reproduction of even the most energetic performances. Of course, you really can't expect to get there with a bookshelf speaker because, among many reasons, you can't transduce the bottom half of the bass. That takes plenty of driver and gobs of juice. For that we simply let you add LF system, as much as you like.

I should point out that having both bandwidth and dynamic headroom in no way lessen the system's ability to play "small and quiet". So those needs are still nicely met. And since, at those levels, the system is barely idling, there's a sense of unrestrained ease.

And, frankly, amp power is cheap so there really is no reason to avoid using it. A 500W N-Core, and absolutely divine hunk o' amp, costs about $100. I can save a few bucks by backing off to the 100 Watter, but why? Then the user would be able to clip his amp during vigorous play; that ain't good. I'd much prefer to have the 'power limit' issue moved up and completely out of play. We use enough power to fry the coils, then prevent that from happening in the processor.

Even though we don't care much about amp power, high sensitivity DOES still mean we use less of it. But the real advantage is cooler coils.

Ok.....
But why is it so hard to find speakers that play well at low volumes? I know the ear is part of the problem here also, but so few speakers seem to be able to reproduce any decent sound at late night, the wife and kids are sleeping, low volume levels.

I guess DSP can help a lot with this stuff. Build a super, low level, software Loudness button, as it were. Tweak the highs and lows. Hmmmm
 
And one more thought.

How did you decide how the speakers would sound? Effectively, with the DSP at your fingertips, you could make them sound like anything you wanted. You could, for instance, do the stereotypical audiophile sound, tilted way up , so you get lots of high end detail. Or the super lush, warm sound, with tons of mids. So how did you decide?
 
While It doesn't entirely surprise me, I'm still amazed at the science behind speaker design. More factors involved than I'd really thought about. Thanks, Pat, for the walkthrough. And, I'm guessing that you're just scratching the surface too. Your passion for audio is so much fun. It's what captivated me, separated you from all the other brands in my mind, and made me such a fan. I'm still saving. I just hope that I can save enough to catch you before you retire.

Do you still engrave your meadowlark logo someplace into your enclosure? I've always loved that. Such an awesome thread.
Thanks so much for the kind words. Yes indeed, speaker design is so darn complex and interactive that it never gets boring. Any two guys will make completely different choices. What matters is that you know what you're trying to do; what's important and how you want it to end up. The reason that most "corporate" speakers kinda suck is that a group of guys with competing ideas make all sorts of compromises with each other during design. Then the bean counters kill it a little more before the production guys de-spec a few more things. To me, that's just great because it leaves the door wide open for small companies with nut jobs at the helm ;-)

We engraved the birdie head logo into the solid wood baffles of our second line of speakers. The first line was silk screened directly to the baffle and rear. Today we're really a custom house making just a few pair each month and things are always changing. Most all are veneered, which does not lend itself to engraving - because you'd be looking into the substrate MDF. So we forgo the logo except for a tiny sticker on the amp plate at the rear.

Since everything now has power on board I'm very tempted to make an illuminated logo, but just haven't worked that out yet.

FWIW - here's the art for the upcoming bass system matched to Black Bird:

kingfisher logo.jpg
 
Ok.....
But why is it so hard to find speakers that play well at low volumes? I know the ear is part of the problem here also, but so few speakers seem to be able to reproduce any decent sound at late night, the wife and kids are sleeping, low volume levels.

I guess DSP can help a lot with this stuff. Build a super, low level, software Loudness button, as it were. Tweak the highs and lows. Hmmmm
Yeah, the Fletcher Munson curves are a bitch. There's no way to make a speaker sound right at both low and high levels. But, yep, DSP solves that problem. You'll find that you can make a Preset for low level and another for head-banging. Another for phono and another for TV.

I'm attempting to acquire the coding acumen needed to devise dynamically modulated Fletcher Munson equalization.
 
And one more thought.

How did you decide how the speakers would sound? Effectively, with the DSP at your fingertips, you could make them sound like anything you wanted. You could, for instance, do the stereotypical audiophile sound, tilted way up , so you get lots of high end detail. Or the super lush, warm sound, with tons of mids. So how did you decide?
Ha! You bet! I recently offered a pair of Kites to one of the biggie review webzines. The editor's reply was "How am I gonna review something that I can make sound any way I like?"

DSP removes a whole group of problems from our attention as designers, thereby leaving us clear to address more interesting issues. You know me: dynamic linearity leads that parade. But there are others, importantly - the believability of vocals. Lack of blur in the bass. Expression of timbre - especially in the bass, where conventional tech has always done a lousy job.

The issue of tonality is kinda personal. You can dial in a flat response at the listening position and it will tear your head off. Personally, I think a few sweet lies - like the one's your lover tells you - are in order. A little extra heft and a slight decline toward the top for sweetness. You will find that, the better your tweeter is, the less decline you'll prefer. You can run Kites pretty relatively hot on top without sensing etch or edge and without any spitty whitening the vocals. On the other hand, Black Bird's tweeter costs 1/6th that of Kite's and requires a little more tender care; still kills it, though.

We usually set up four Presets with slight differences in midrange presence and let the user tell us which sounds nicest to him. Then we kill the remaining Presets and configure that one for other variations. Room EQ, etc. So the adjustability has real value worth paying for.
 
So, you could in theory, mimic the sound of any other speaker. If I wanted my new Kites to sound like my old Shearwaters, the DSP would allow for that? I get that the driver's are far better, but would this be possible?

We could create a file with no highs or lows , and recreate a Bose sound? :)
 
Back
Top