Dutch & Dutch 15C Discussion

[illusonic]

This thread is about the 15C. It is a True Monitor Loudspeaker fully capable of reproducing a full spectrum 16 hz to 20 kilohertz recording. And at levels across the spectrum higher than 105dB. So why do you even bring up a boutique midrange that requires a limited bandwidth and limited SPL output as a comparison? The entire reason why a 15 inch midrange was chosen was to have no limitations in the lower bass, midrange area. The distortion levels are actually difficult to measure. Radiation patterns between the midrange driver and the horn are as matched as can be possible. Again, planned.

I went through the simulation to demonstrate that the 15 inch driver choice has reasons behind it, and they are valid. The 15C is a part of a design philosophy. The Dutch & Dutch guys like a 2-way. The Bleisma could be used in a 4-way preserving full hemispherical output from each driver across it's range. I'm sure that you know this.

The 3" dome supports most of the midrange requirements in the link postend by @Sigberg Audio. That's why I brought it as an example. A 15" driver will struggle with some of these requirements. It's designed for a different purpose.

The discussion started with the following posts:

In particular the fact that using monstrous 15" driver for the midrange has no penalty for this audio band reproduction fidelity,

It either has the bandwidth to handle midrange with low distortion or doesn't. If it does, then the required time domain response follows.

THD vs. level would be interesting to see for large midrange drivers. Especially at low SPL.

Now you keep writing about maximum SPL capabilities. It's trivial that a 15" has much higher maximum SPL than a 3". And it's also obvious that the 15C was designed for high SPL. But that's just not the point of the discussion.

It's a monitor. 1.5 to 2 metre is quite normal. What is interesting about this design is that it is perfectly capable of outputting 105 db at 2 metres and then have headroom remaining. The bass design request was for 105db at 16 hertz. That was exceeded by quite a large margin.

When sitting at 1.5-2m distance it may happen at times that also low playback level is desired. This brings up the question if midrange could be compromised at such levels. Potential reasons for such compromises can be found at the end of my last post.

 

[Sal1950]

Slightly off topic in this thread perhaps, but I have notified the author of the article about your comments, let's see if he responds. I don't know enough about driver design to comment.

That would be good, he had already reserved post #2 for "detailed spec's" but hasn't populated it in 6 months now. Isn't this a finished design?
This really is quite a awesome speaker but I think we already knew what could be accomplished in freq extension, spl, etc; by using a bunch of large bass drivers around a good "horn" top, with a active dsp crossover plus a lot of power. I don't really mean to throw dirt on the design, it is a herculean effort at building a massive monitor. But I do wonder if it will really get us any closer to that "transparent to the source" dream?. In my mind I tend to doubt it, I may be wrong but when that day comes I tend to believe it will somehow use a minimum of drivers, one would be ideal if it could be designed. This speaker is really nothing much more than a JBL M2 with a few more bass drivers to obtain more SPL with lower distortion, but how many have rooms large enough to use that? And does the number of drivers cause more confusion in their integration possibly blurring the resulting output over a "less is more" approach?
Just sayin

 

[mwmkravchenko]

Which means about 16um peak for 75dB SPL. Now I would start to worry if the mechanics of the 15" speaker are optimized for such low excursion. I'm thinking about effects like stick–slip friction in suspension and viscoelastic hysteresis. If there are such kind of issues, near-field measurements of e.g. THD/IMD vs. level and time domain analysis of sine bursts could probably reveal them.

I didn't reply to this because it is not an issue. And has never been an issue. Sound is air pressure differential. How you produce that is what matters. With as little added variation to the incoming signal as possible is the true goal. You are right, it's about 6 microns of movement each way to get your desired SPL. The great think is you have absolute maximum linearity in a drivers suspension at these movements. That is the reason you do Parameter measurement at small signal levels in the first place.

 

[Snickers-is]

I found some of the comments in the article puzzling. Low carbon steel is fully saturated at about 2.2 Teslas. That is not across the top plate pole piece gap. That is in the steel itself.

The article does not claim that saturation is measured in the air gap. Saturation is defined as the point where the permeability approaches that of vacuum. In air, that is already the case.

That only gets accomplished in Compression drivers. And perhaps very high efficiency wideband drivers.

It can be done in any type of driver, if the budget allows for it.

For most purposes a fully saturated top plate and pole is actually not useful. The more powerful the motor the greater the self damping effect, like shocks on a spring system in an automobile. The higher the motor strength the more rapid the damping properties of the motor.

The simulations I have made over the years shows a very clear link between some types of intermodulation distortion and the level of saturation. Another factor is the relative proportion of the fixed and the variable magnetic field. Multi layer designs have wide gaps and strong variable fields, and they tend to perform worse on IMD.

The damping effect only happens as a direct result of deviantions in the permanent field to begin with, and they are really ineffective at low frequencies as the ability to transfer energy is exponential degrading towards lower frequencies.

While increased current density will transfer more energy, and be more effective at this than a coil with lower current density, this will also be proportional to the scale of the variable field, so you have not solved anything.

To put it very short and simple:
If you are far from saturation, you add a layer of complex reactive components directly in the signal path.

Also the choice of praising Kevlar. I find that they have nasty breakup resonances within the drivers passband.

The article is not praising kevlar, and it does not recommend using a woven diaphragm in its break up region. I also think you are stretching the definition of passband if you include the break up region.

 

[bmc0]

Now I would start to worry if the mechanics of the 15" speaker are optimized for such low excursion. I'm thinking about effects like stick–slip friction in suspension and viscoelastic hysteresis. If there are such kind of issues, near-field measurements of e.g. THD/IMD vs. level and time domain analysis of sine bursts could probably reveal them.

As I said before, I've never seen evidence of such problems in acoustic measurements. I just did some nearfield harmonic distortion measurements of a 12" PA woofer (Eminence Deltalite II 2512) at 1kHz and the equivalent of 75dBSPL@1m to 25dBSPL@1m in 10dB steps. 128k FFT with 8 coherent averages to suppress noise. The remaining noise and spurious junk is below the absolute threshold of hearing at 1 meter (-140dBFS corresponds to approx. -17dBSPL@1m):

 

[Duke]

This speaker is really nothing much more than a JBL M2 with a few more bass drivers...

If we paint with broad strokes, then perhaps so.

However the devil is always in the details, and in this instance the details differ devilishly.

In my wicked opinion.

 

[illusonic]

As I said before, I've never seen evidence of such problems in acoustic measurements. I just did some nearfield harmonic distortion measurements of a 12" PA woofer (Eminence Deltalite II 2512) at 1kHz and the equivalent of 75dBSPL@1m to 25dBSPL@1m in 10dB steps. 128k FFT with 8 coherent averages to suppress noise. The remaining noise and spurious junk is below the absolute threshold of hearing at 1 meter (-140dBFS corresponds to approx. -17dBSPL@1m):
View attachment 541701 View attachment 541702 View attachment 541703 View attachment 541704 View attachment 541705 View attachment 541706

Thanks for your effort! Indeed results show no issue with HD at very low levels. Does this mean dyamic range of a driver always increases with maximum SPL? So one can just pick the highest SPL option? Somehow hard to imagine.

 

[Salt]

Thanks for your effort! Indeed results show no issue with HD at very low levels. Does this mean dyamic range of a driver always increases with maximum SPL? So one can just pick the highest SPL option? Somehow hard to imagine.

You do not really understand how transducers work an measure.

 

[illusonic]

The damping effect only happens as a direct result of deviantions in the permanent field to begin with, and they are really ineffective at low frequencies as the ability to transfer energy is exponential degrading towards lower frequencies.

When I vary the size or strength of the magnet in a speaker, it results in different levels of electrical damping at the resonance frequency. This is very effective to get the desired Qts.

 

[Snickers-is]

When I vary the size or strength of the magnet in a speaker, it results in different levels of electrical damping at the resonance frequency. This is very effective to get the desired Qts.

Oh, absolutely! But that is tuning of T/S parameters. It will not directly affect distortion.

The topic discussed is modulation of flux inside the steel parts of the motor, and the distortion it leads to.

 

[illusonic]

The topic discussed is modulation of flux inside the steel parts of the motor, and the distortion it leads to.

You mean variation of VC inductance over excursion caused by non saturated steel?
Edit: Seems you refer to BL(i) and Le(i) rather than Le(x) although the latter usually causes higher distortion.

 

[Sal1950]

If we paint with broad strokes, then perhaps so.

Yes, very broad strokes! I'm just making the point that the same basic tech approach is used to accomplish the desired goal. That goal is a speaker with good linearity that's also capable of very high SPLs while retaining reasonably low distortion. We do know exactly what the M2 is, at $30,000 a pair

• Wide radiation pattern that is not just controlled but constant through a very wide bandwidth
• Very neutral on-axis response (with some exception to the midrange resonances discussed above)
• Extremely high output capability and flat down to about 30Hz (with in-room response likely extending lower)
• Practically no compression (and I imagine nearly anyone’s ears would give out before compression was even remotely an issue)
• https://www.erinsaudiocorner.com/loudspeakers/jbl_m2
• ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Still no solid measured info for the 15c but I'm sure it will be fantastic!
BUT 5x more at $150,000 a pair better ???
The JBL M2 is beginning to look like a steal !

 

[Digital_Thor]

Further... When comparing them to the M2. The difference in one being active and another passive, I also think it's worth mentioning that amplification, dacs and streaming is quite cheap now, especially since this site in particular, has shown that it does not need to be crazy expensive to be really good.

 

[ppierre]

When comparing them to the M2. The difference in one being active and another passive

I think both have external amplifier and active crossover.

 

[Digital_Thor]

I think both have external amplifier and active crossover.

Really? I would definitely have thought that the Dutch was a complete package, like it's siblings?
But then at least, it's all included in the price, right?

 

[Sal1950]

Really? I would definitely have thought that the Dutch was a complete package, like it's siblings?
But then at least, it's all included in the price, right?

The only real difference is the D&D speakers contains DSP crossover and amplification inside the speaker itself.
The M2 requires outboard electronics.
The price I quoted includes the Crown pro series amps & DSP crossover components.
There is also "audiophile" level packaging using Mark Levinson components and others for those preferring a "roll your own" approach.

 

[Snickers-is]

You mean variation of VC inductance over excursion caused by non saturated steel?
Edit: Seems you refer to BL(i) and Le(i) rather than Le(x) although the latter usually causes higher distortion.

Well, all of those are parts of the equation, but if you run more complex sims like Bl(x,i) as a function of voltage input, you map x and i together for different frequencies (not much i is required for high x at low frequencies, but not much x will result from high i at high frequencies), you add in the effect of Le(x,i) (to get the effect of Le) on top, and you also check the polar properties at each point, you will quickly find that the amount of distortion caused by layers of complex reactive effects is really significant.

 

[Waxx]

The only real difference is the D&D speakers contains DSP crossover and amplification inside the speaker itself.
The M2 requires outboard electronics.
The price I quoted includes the Crown pro series amps & DSP crossover components.
There is also "audiophile" level packaging using Mark Levinson components and others for those preferring a "roll your own" approach.

Wel, the most "roll your own" approaches that i've seen where hypex or purify amps and MiniDSP or Trinitov DSP's to do the filters, not the Mark Levinson amps (altough they are certainly not bad, but a relic from the past for many) that does not have the power needed for this speakers. I've also seen a setup with Powersoft amps and (Cosine) dsp in an HT setting.

 

[Soniclife]

The only real difference is the D&D speakers contains DSP crossover and amplification inside the speaker itself.
The M2 requires outboard electronics.

The 15C uses external amps and crossover, that will be part of the package.

The real difference is the dispersion, they have very different approaches, look at the 8C review for what I assume the 15 will look like, though the width of the beam may be different. I'd assume a general improvement in graph prettyness as well. The different dispersion will lead to very different in room sound.

 

[Sal1950]

Wel, the most "roll your own" approaches that i've seen where hypex or purify amps and MiniDSP or Trinitov DSP's to do the filters, not the Mark Levinson amps (altough they are certainly not bad, but a relic from the past for many) that does not have the power needed for this speakers. I've also seen a setup with Powersoft amps and (Cosine) dsp in an HT setting.

Yes but your seeing it thru 2026 eyes. Back in 2013 there was a factory upgrade package using ML electronics, now obsolete I believe.
A very popular build I hear of now uses Benchmark amps and various DSP, using the factory crossover specs, or not.

The 15C uses external amps and crossover, that will be part of the package.

I stand corrected there.