Dynaudio Special 40 - WTF!

[thewas]

Nope. Look at the crossover, its a 1st order with impedance compensation.

1st electric order, we are talking about total acoustic order.

 

[RobS]

Then Vandersteen and Thiels are the only ones I can think of that do first order acoustic slopes.

 

[RobS]

But the other guy is incorrect in that the SP40s are a "typical 2-way, 4th order step response". If you look at the picture of the crossover, it's enough to know its not 4th order. And looking at a step response is not decisive in determining the crossover topology.

 

[thewas]

AND, the linear-phase result is only achievable/measurable at a single point in space.

Also correct, although with a coaxial driver it could have been approximated more.

 

[YSDR]

Nope. Look at the crossover, its a 1st order with impedance compensation.

The reason for the step response is they are linearizing the phase response with precision filters or they designed the woofer to match the bode plot of the filter they already designed.

Nope. A first order crossover step response never looks like the S40 have. 1st order xo not even have one single sharp negative swing in the step at any point in space, even if it's phase aligned with additional elements in the crossover.
Here is a real 1st order xo step response:

https://www.stereophile.com/content/thiel-cs2-2-loudspeaker-measurements

 

[ROOSKIE]

But the other guy is incorrect in that the SP40s are a "typical 2-way, 4th order step response". If you look at the picture of the crossover, it's enough to know its not 4th order. And looking at a step response is not decisive in determining the crossover topology.

Yes but speaking only of actual acoustic roll off. Is it not possible to achieve 4th order acoustic roll-off or close to it, with a filter that is 1st electrically. Granted the natural driver roll of would have to be very, very optimal. I do believe that tweeter is claimed to have such properties.

It's insanity. Tweeters simply cannot absorb or dissipate the power/heat at those frequencies and at any decent levels. Not just dynamic compression you'll be dealing with, but cooked/vaporized VCs.

Well, there are quite a few passive designs that cross 25-29mm tweets below 2k. I mean it is not uncommon at all. (not counting compression drivers which easily cross here and lower)
Even more actives.
I even have made two as a hobbyist and they sound/measure great. One is crossed at 1400hz no problem even at high volumes. It is a 2way with a 6.5" and the woofer gives up first (when not high passed to a sub).

 

[YSDR]

But the other guy is incorrect in that the SP40s are a "typical 2-way, 4th order step response". If you look at the picture of the crossover, it's enough to know its not 4th order. And looking at a step response is not decisive in determining the crossover topology.

The electrical order of a crossover doesn't mean too much in itself. For example, you can have a 1st order electrical filter that produces 4th order acoustical slopes if combined so with the raw inbox responses of the drivers.

 

[ctrl]

But the other guy is incorrect in that the SP40s are a "typical 2-way, 4th order step response". If you look at the picture of the crossover, it's enough to know its not 4th order. And looking at a step response is not decisive in determining the crossover topology.

To add a little to what the others have already said....

If you look at the step response, you can see that both drivers are connected with the same polarity.

Source: Stereophile
If we assume that a common crossover is used (excluding exotic mixtures of different filter slopes), then it is most likely a third order crossover, which would be unusual, since normally polarity reversed drivers are used there because of the better group delay, or fourth order.
With a second order crossover, polarity reversed drivers must be used.
Also the time gap of the two peaks fits quite well for a third or fourth order crossover at 2kHz.

 

[ctrl]

I have to take Danny's side a little bit. The speaker is already doomed to failure by design.

If you place the tweeter so that the distance to three cabinet edges is almost identical and then don't adjust the crossover frequency to minimize the negative effect of edge diffraction, then all is lost - in terms of directivity.
In the DIY Directivia thread we also briefly addressed this case (the initial post is that).

The crossover frequency of 2kHz with the original crossover, can not iron out much.
If the crossover frequency is then lowered to 1.2kHz in Danny's version, the matter is only slightly worse.

To show this, I took from Danny's measurements the 40° frequency response (Danny did not say what is shown in the picture, 40° is my guess) ...

... normalized it to the on-axis FR and compare this to the normalized FR of 30° and 40° by the stereophile (the diagram is pretty crap and very confusing, did my best to extract the frequency responses and scale them similarly):

The curves are not directly comparable (the sources are too inaccurate for that), but show that the dispersion problem around 3kHz also exists with the original crossover.

 

[RobS]

Directivity is fine to a certain extent, but dispersion is always going to be more desirable. Don't want to clamp your head in a vice to listen.

 

[amirm]

Would be cool to see, but you'd probably be without them for a while, risk damage in shipping (probably the most likely) and I doubt Danny's going to want to risk being put underground by the NFS.

I have indeed expressed willingness to test this mod. I even offered to pay to ship it back. Seeing how Danny makes his entire case using measurements, it is so odd that he doesn't want me to verify his work.

 

[RobS]

He probably doesn't want you to reveal his "upgrades" are a waste of money and make the speaker worse.

 

[amirm]

He probably doesn't want you to reveal his "upgrades" are a waste of money and make the speaker worse.

It is the only logical conclusion. Seeing how I am donating the time and effort to test his product with no profit motive, whereas he has one, it doesn't put his answer and conduct in good light.

 

[Kachda]

between @amirm 's japanese woman and @RobS hot chic with a beer cup, I can hardly read the words.

 

[puppet]

It's too bad the fella, Danny, has that mindset. If I were in his shoes, I'd jump at the chance to have my work verified by a third party.

Yeah, I know, why should he. He's got no reason to lay himself open. ... but on the flip side, it would go a long way to "partnering up" with other audio pro's. In the end, I'd think it would be a better approach to increasing business. Too much animosity in the world today. What a drag.

 

[Rick Sykora]

No before/after distortion measurement or check on phase alignment between the drivers?

These are essential checks to see if the tweeter is being overextended and how well the new crossover integrated the drivers.

How about a before/after measurement on those tube connectors too? Does he mod his electronics so the other end of the speaker cables have the alleged improvement he claims? Any sign of a listening test that was supposedly going to prove what measurements cannot?

Too many questions without any substantive answers.

 

[Sugarbubble]

Sigh. Wake up folks. That guy is a charlatan. The sf40s are fine in the listening environments that I have heard them in. In each case (3) they were a stereo pair used in a living room environment. Does anyone other than Erin and the manufacturer measure them in real world settings? Are they great? No. Are they fine? Yep. too many people have distorted value propositions here in which they wish every component to be dirt cheap and have the best measurements. People I hang out with have different priorities.

 

[Pikas]

Is there a standardized rig set up for journal publication?

 

[goldenears]

too many people have distorted value propositions here in which they wish every component to be dirt cheap and have the best measurements. People I hang out with have different priorities.

The kind of people who might buy expensive upgrades for speakers could possibly have those same priorities, no?

 

[Lavawood]

These speakers 3 feet into the room sound incredible. I understand the subjective measurements to a degree but in your listening room, from your listening position, I think you’d be hard pressed to find better bang for the buck. Anyone who owns these, you did good. Anyone thinking about it, don’t let this discussion deter you, they’re wonderful. Hooked up to a rogue Cronus magnum in a 10’-15’ room is heaven.