• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Smart Technology for Performance Loudspeaker/Driver Manufacturers to Implement to Reduce Distortion - No Licensing Required!

mike7877

Major Contributor
Joined
Aug 5, 2021
Messages
1,051
Likes
262
"...standard steel pole pieces and front plates in speaker magnets suffer from nonlinear magnetic characteristics. As current fluctuates through the voice coil, the ... steel lags behind the varying [electric] magnetic field (hysteresis), creating unwanted eddy currents ... causing significant third-harmonic distortion."

"...replacing the standard steel areas that run concentric with the voice coil with..."

"...a specialized composite—typically made of powdered pure iron, resin, and PTFE."

"...prevents eddy currents from forming in the first place, while its high permeability maintains ... magnetic flux."

"...allows... :
- Lower third harmonic distortion by 10dB to 15dB in the 100Hz to 3kHz range.
- cleaner reproduction of complex sounds, [most notably and perceptibly improved are the] human voice and piano."


I reduced the amount of words by half, keeping everything pertinent to save you a minute of your life :)


So... This material... It's been around and its properties have been known for many decades now, so a reasonable person might think "if it's free to implement [excluding {inexpensive} materials], and it's so good... everyone is using it already, right?"

But you'd be wrong. One company has been using it for decades, and from what I've been able to gather they were very forthcoming with details around launch time as well as after. But still... they are the only company!!!

"How could this be?"

Well, I have a couple ideas why it's not used in the majority of speakers:
- most speakers' THD wouldn't be improved by 10-15dB
- exceptional fidelity is only a selling point for a minority of speakers

"But there are many high performance speakers sold. Surely they could benefit?"

Probably a good number of them, yes. There are a lot of companies that sell speakers that are supposed to be high fidelity and aren't. The companies that would benefit from this technology most are ones that
- sell most of their speakers in the $1500-$8000 range

The more expensive speaker companies could benefit too, but they're often more focused on making a bunch of money than being honest (if they were honest, they wouldn't be selling $7,000 of parts with a couple hundred man hours in R&D for $150,000).

The companies that sell high quality with fair profit margins (90% of their sales between $1,500-$8,000USD), their designs, with a bit of refinement here and there (eg. make sure passive crossovers are air core and properly aligned, upper-mid tier capacitors), could be substantially improved with this technology on their woofers and midranges.


So which company is monopolizing this technology? ATC. They're in the UK mostly.
They use it in their highest end professional series. It's in three different sized woofers (with the same motor), and one midwoofer (also the same motor).

The motor? It's maxed out with double stacked ceramic magnets totaling ~400 ounces, mounted on a THICK cast aluminum frame. More details:
- rectangular CCA wire
- underhung 8mm voice coil in 20mm gap
- active individually tuned 4th order crossovers
- CLD driver (layered driver with middle damping layer to smooth out frequency response as directionality increases

there are more things about the woofers that make them awesome like balanced suspension, an xmech of 4x the xmax, extremely linear magnetic field through the 20mm gap), but that's not important - showing the caliber of woofer this belongs in was the point.

The three woofers are 9, 12, and 15 inches, and the midwoofer is 150mm (actually the cone+1/3 surround measures 140mm or so)


For the last multiple decades, ATC has only done small refinements to their designs. And that's fine - their stuff is great. The material works best at reducing distortion between 100Hz and 3,000Hz, but because ATC's biggest sellers by far are the speakers with the 9, 12, and 15 inch woofers, and those woofers are crossed at 380Hz 4th order, that means only 280Hz are really being optimized for low distortion by this tech. Like 90% of the drivers made with this only hit 380Hz. Personally I have the 150mm which is crossed at 2.1kHz. I enjoy it a lot, but I think it's really sad that it's the only model speaker like it. ATC doesn't seem to be working with DSP, and in the last 5-7 years, DSP has come a long way, so I'm raising awareness.

Manufacturers of quality speakers/drivers should use this in their future designs! It doesn't take much to implement, and starting with good designs, maybe you only need to do a couple things to squeeze the max performance out of things!
 
Are we sure that this is actually underused? Purifi has been pretty forthright about their efforts to reduce hysteresis distortion. I would be surprised if the big names casually missed an easy, legitimate upgrade to get 10dB distortion out of the midrange.
 
ATC runs their SM75-150 dome midrange from 380Hz-3000Hz in their speakers. Yet the SM75-150 doesn't use this hysterisis-reducing compound. Why is that?
 
Are we sure that this is actually underused? Purifi has been pretty forthright about their efforts to reduce hysteresis distortion. I would be surprised if the big names casually missed an easy, legitimate upgrade to get 10dB distortion out of the midrange.
The Starship Enterprise has been using it to keep the Dilithium crystals in check, since like, ever. Data for this material, and it's effects is tough to find. It's magnetic, Jim - but not as we know it...

Cheers
 
Because copper shorting rings work just fine for both over and underhung motors, where this is only useful for underhung. Why use a special tech that doesn't work for all your motors?
ATC runs their SM75-150 dome midrange from 380Hz-3000Hz in their speakers. Yet the SM75-150 doesn't use this hysterisis-reducing compound. Why is that?
It actually does.
 
Where are the measurements?
ATC has a white paper on it floating around. It's basically doing the same thing as copper shorting rings albeit with a different mechanism.
 
Because copper shorting rings work just fine for both over and underhung motors, where this is only useful for underhung. Why use a special tech that doesn't work for all your motors?

It actually does.
No, it doesn't. I own nine of them, and I can assure you they are plain steel. They also don't have shorting rings.
 
Are we sure that this is actually underused? Purifi has been pretty forthright about their efforts to reduce hysteresis distortion. I would be surprised if the big names casually missed an easy, legitimate upgrade to get 10dB distortion out of the midrange.

Well the rest of the driver has to be good enough to see the improvements to begin with. So first, it can only be used to improve underhung designs. Next, if third harmonic distortion isn't already -60dB from 100Hz to 3000Hz (and better), the improvement would be completely masked.

Also they could be put off by weight - significant weight is added... The basket needs to be cast, and the magnet needs to be about 10kg (for any size woofer, right down to 5" lol). You might be able to use neodymium instead of ceramic, but I don't know all the properties of the special aforementioned material - neo might be too small and strong for the field to be made uniform by the time it reaches the gap.

A lot of companies, all their woofers are overhung. Overhung can be great up to a point, but they have a limit. If all of a company's products are ported and have xmax of 25mm from overhung (ie. slowly increasing distortion past 25mm), and they release one with an xmax of 10mm and no port, with distortion increasing significantly once more than a few percent past 10mm, I don't know... Decision makers might say "it's too different" from their other products, unless they want to create a new range with a few different models in it, possibly.

ATC runs their SM75-150 dome midrange from 380Hz-3000Hz in their speakers. Yet the SM75-150 doesn't use this hysterisis-reducing compound. Why is that?

The reasons I can imagine they're not implementing this on their domes
- negligible performance increase
- space consideration
- efficiency consideration

ATC designed their new(ish) 1" dome tweeter (SH25-76 & SH25-76S) after their 3" dome midrange. It's basically the same thing but shrunk down - most notably and relevant for this topic is dual suspension and being underhung.

The SH25-76 is about 3dB less efficient than the SH25-76S, and I believe this is entirely down to the stronger magnet on the "S-Spec" tweeter. ATC says that the 3.0 mm magnetic gap of the S-Spec tweeter is completely saturated, meaning saturated through the area that the voice coil moves. One way that the material mentioned in the OP reduces distortion, is by maintaining uniform strength of the magnetic field through the gap. I don't know all the ways or how much this property is generally responsible for the total distortion reduction of the material, but my educated guess is it's the vast majority. I think the other part of it is reducing reactance.

So the S-Spec tweeter is used in the higher end professional speakers, and the standard spec is used in their Hi-Fi line.

-2dB points of the SH25-76 are 720Hz and 17kHz (20kHz -3dB, 22kHz -6dB)
-2dB points of the SH25-76S are 1.5kHz and 20kHz (22kHz -3dB, 26kHz -6dB)

I believe the SH25-76S tweeter has a higher low frequency cutoff from the magnet's strength (ie. strength causing low Q)

Having a normal Q (of probably ~0.7) and being flat to 700Hz makes the non-S-Spec tweeter very easy to keep phase through driver transitions of passive crossovers.

A lot of ATC's non studio speakers can be bought passive or active.

S-Spec tweeters are used in ATC's highest performance (and cost) speakers, which are usually required to be much more active over their life cycle. It's only used in active designs with one exception, there is a passive variant of the SCM20 Pro Mk2. It's 2 way and crossed at 2.1kHz 3rd order Butterworth, with all other speakers [Studio or HiFi] being 4th order crossed at 3.5kHz (except the other SCM20 Pro Mk2, the active one, it's second order, 2.5kHz...)


I'm quite sure that the S-Spec dome midrange (used in all the studio versions) uses roughly 10kg of ceramic magnets - around same amount as the woofer. The woofer's magnetic gap and voice coil are both double the midrange's though - the 20mm gap and 8mm voice coil of the woofer is approximately double that of the mid.
The next two figures are possibly unreliable memories, but I'm pretty sure BI of the woofers are 8, and the midrange (S-Spec): 16.

I think having >20 pounds of ceramic for a measly 3" driver would saturate the magnetic gap, causing the uniform field, and, so, the benefit of the linear magnetic material mentioned in the OP would be minimal.


For woofers, it would most often not be desirable to have a saturated magnetic gap. With a saturated magnetic gap, a woofer of normal efficiency is going to have a Q that's way too low to be useful for any sealed designs. Since the material allows for a uniform spread, and you could actually use it to tune the driver's Q by varying the width of the gap!

To sum it up for tee ell dee are ers , I think ATC didn't use the material in the S-Spec dome because its magnetic gap was already saturated so the distortion wouldn't be meaningfully lowered by it. And then they didn't put it in their standard dome because, possibly, there wasn't space for it or there would be too much efficiency loss, or Q would end up rising too high.
Or maybe it's intentionally worse for product differentiation! (I hope not lol)
 
Where are the measurements?

Measurements of what, specifically?

ATC has done measurements of it and concluded that in their designs, it is responsible for a 10-15dB reduction of 3rd harmonic distortion.

I think I mentioned the woofer design. I'll go over it, possibly a little further this time:
Underhung
Cast frame
Balanced suspension
Vented
CLD diaphragm
2 layer 3" 8mm voice coil of edge wound rectangular CCA wire on Kapton
10kg ceramic magnets for 20mm gap

And the CLD method is very refined, all design aspects done correctly.
(when sold, is fed by proper high power high current low noise low distortion class AB amplification (first third class A), and passive use only properly arranged and oriented air core inductors and high quality capacitors and everything else right and good including connectors and connections and and and)


This is the caliber of speaker which is able to actually realise this 10-15dB improvement. ATC says maintains at least -70dB through the range - what I remember of the chart accompanying it: showed the average to be ~-80dB, maybe 1-2dB better, with the distortion falling starting from -70dB @100hz, reaching -80dB by 500-600Hz, with a full range of -70dB to almost -90dB between 100Hz to 3000Hz (not a constant slope down - the highest half is mostly flat)
 
(sorry for the 20 day delay - I didn't realise I posted the thread)
 
Measurements of what, specifically?

ATC has done measurements of it and concluded that in their designs, it is responsible for a 10-15dB reduction of 3rd harmonic distortion.
In that case, measurements of the 10-15dB reduction of 3rd harmonic distortion
 
In that case, measurements of the 10-15dB reduction of 3rd harmonic distortion
There's an AES paper about it and ATC put out a whitepaper a long while back. Not sure if I'll be able to find either.

EDIT: Found the AES paper. Attached below.

EDIT 2: The whitepaper: https://atc.audio/wp-content/uploads/2014/10/ATC-CORP-BROCHURE-FINAL-10.pdf


TLDR: Achieves similar things as shorting rings (i.e., lower distortion from hysteresis and eddy currents) but essentially in a backwards approach (shorting rings do as they describe, i.e. short out the eddy currents, SLMM basically acts as an electrical insulator so the currents don't manage to reach the voice coil from the metalwork).
 

Attachments

Last edited:
Back
Top Bottom