New Project - tower 3-way with twin 8s

There is a good description of LTI and response correction in the paper Correcting Loudspeaker Transient Response from David Gunness at Fulcrum Acoustics
Whitepapers | Fulcrum Acoustic



I can't link the paper directly from a phone.

For general IIR parametric EQ I would just measure the driver on the intended baffle and look at the effect of on axis EQ on the off axis responses in vituix. If you can see negative consequences from the on axis EQ then tone it down.

While it is good to have lot of knowledge on how things actually work it is very liberating to tinker with VituixCAD and see the stuff happen even without knowing whats behind, as fluid describes :) For example it is easy to see that certain properties of the response cannot be fixed with EQ and others can be simply by looking a polar graph, or the DI, while tuning an EQ. You'll notice a parametric doesn't have any effect on DI unless EQ is on single driver on crossover region.

I mean, just by tinkering and observing what happens one is able to tune in a very good response, if the physical construct allows very good response. By good I mean smooth DI, in which case the on axis and power response can be smoothed with the EQ. I suspect that with enough hours behind one is able to conect the graphs mentally to the sound so that certain things can be sacrificedforothers. For example a smooth DI seems to yield very good sound, but there might be some reasons to sacrifice the smoothness for even better perceived sound. Anyway, very powerful helper it is for thinking and tweaking. Anyone can come up with very good crossover for a system just by observing, without knowing what phenomenon there is behind all of it.

Whitepapers | Fulcrum Acoustic

My latest 3-way uses Hypex FA123, drivers are SB26STAC, 4" Audax HM100Z0 and 2x SS21W (Discovery pulp cone) in closed box. Sorry I don't have high quality measurements, it was wintertime and speakers are too heavy to be carried outdoors...

Speaker design and response tuning is always a compromise and listening is important too. By looking at sims and measurements alone one easily goes to too little details and faults, that don't have any audibility.

DSP units make it easy to do experiments, how gross an EQ setting has to be to be audible. High Q 5dB is hard to hear, but low Q 0.5dB might be heard. Likewise uneven directivity at M/T crossover range gets "averaged" in listening window (CTA2034) and that is real too! Vertical axis problems can be worse, but response can be tuned neutral to listening spot.

Anyway, 3-way with acoustic LR2 or Duelund topology easily gives real hifi sound!

https://www.diyaudio.com/forums/multi-way/321711-avalanche-as1-modernization-2.html#post6521760


Thanks a lot for the suggestions @fluid, @tmuikku, and @Juhazi :)
I am definitely learning a lot from you all. I will definitely go and read up the articles mentioned in your posts and tinker around with VituixCAD more and understand what aspects of DSP are more important in speaker design and more importantly, in the overall perceived sound.
In my daily job, I have to use DSP on a regular basis in the design and testing of wireless chipsets, hence all these curious questions regarding the DSP related aspects. Sometimes curiosity gets the better of me.. :D
I am now beginning to understand more about this topic and how it is different from DSP applications in my domain from all your suggestions.

Thanks a lot, again. :):)
 
Vineethkumar – I have been following your thread with interest. I considered a waveguide, and if I did go that route, I would have used the Augerpro design. But I am staying with a flat baffle for this project.

As tmuikku said, VituixCad is a very powerful tool to do exploratory simulations, to tinker and try things out. One thing I have found is the importance of the diffraction hump (there may be a more proper name for this). When crossing from a mid to a tweeter, the size and location of this hump is important. If the crossover is within the hump region, it will be very difficult to achieve both a good on-axis and good power response… in other words, a smooth DI. A wider baffle causes the hump to be lower in frequency. In my graphic I show a 350 mm x 250 mm baffle with a 5" driver in the middle, to illustrate the hump.

The greater dynamics and higher SPL_max of a 6" SB17CAC35-4 midrange seems a better match to two (21mm P-P) SB23MFCL45-8 woofers
Perhaps this is true. I wanted plenty of headroom in the bass. The SB15CAC crossed at 300 Hz requires minimal BSC, as it is radiating mostly in 2-pi. The woofers on the other hand are radiating into 4-pi. They also have a 6 dB bass eq Linkwitz Transform. Both of these factors reduce headroom in the bass. At 40 Hz, a single SB23MFCL45-8 becomes displacement limited at 99 dB with 77 W signal. A pair would be 105 dB. I think the SB15CAC30 will be just running out of steam at 105 dB, so it may work out fine. We shall see…

j.
 

Attachments

  • Presentation1.png
    Presentation1.png
    458 KB · Views: 470
I strongly recommend using a dedicated midrange instead of a midwoofer like SB15CAC. High excursion is not needed and the enormous surround roll of CAC will excite IMD. Many midranges use inverted roll instead for a good reason. Mids have better efficiency too, and highpass can be adjusted based on distortion tests in the box (easy again with DSP).

3-way design should be addressed to minimize midrange IMD/Doppler and to push MT xo higher than in typical 2-ways - above baffle step and difraction range. Using 2nd order acoustic slopes we also minimize excess group delay (phase rotation) which will as well make eg. piano to sound much better, almost like a very good single fullrange speaker.
To achieve this I think a 4-5" dedicated midrange driver is the most important and critical choice of the design process. Baffle shape might come second and xo third.

Here is SB15CAC, and notice how the "solid" radiating area Sd is actually pretty small (~95mm diameter). SB gives speaker number by frame's outer diameter 150mm)! This is why we cant just look at model number from different manufacturers!

SB Acoustics SB15CAC30-4 5" ceramic woofer- 4 ohms

sb15cac30-4-mech.jpg


sb15cac30-4.jpg
audax-hm130z0-midrange-525-hda-cone-8-ohm.jpg


This info from Purifi was new to me Distortion, The Sound That Dare Not Speak Its Name - PURIFI

"Surround Radiation Distortion

The surround is generally considered part of the suspension, and as such only its contribution to Kms(x) is noted. Its contribution to sound output is rarely noted. In modern (i.e. long-stroke) drive units the surround can easily make up 20% of the radiating surface. It would be optimistic in the extreme to expect that a piece of deforming rubber will end up radiating undistorted sound. Indeed, it does not – its distortion contribution exceeds that of the cone by orders of magnitude. Again the distortion is second order in nature and most obvious at low frequencies. But again that introduces intermodulation distortion affecting the entire frequency range of the driver. SRD is the second reason why large diameter, short stroke drivers have a leg-up. The surround simply takes up a smaller percentage of the moving area. Clearly though, that is only half a solution. The real solution lies in finding a design whose acoustical output is distortion free."

And more here Some Speaker Problems That Needed Solving - PURIFI
 
Last edited:
  • Like
Reactions: 1 user
I appreciate your recommendation, Juhazi. You bring up some good points.

This is not a state-of-the-art high performance project. If it were, I would have considered the Purifi driver, or satori MR13, or some other dedicated midrange. But given the budget of $500 per side for drivers, I think the SB15CAC30 is a good compromise. I was very impressed by the performance and sound quality of the larger SB17CAC35 when used as a midrange driver.

In any case, the drivers are already purchased.

j.
 
I am using the SB23MFCL45-8.

Given the global supply chain issues, I ordered the drivers much earlier than I need them. They arrived yesterday.

I was looking at these same drivers to pair with the textream 6 1/2 mid and textream tweater both of which I have. I have them currently in a test box with one 8" silver flute and they sound good but they are lacking a good strong low end. So I was debating between these and the Dayton 7" Epique E180HE-44 but haven't heard either of these drivers in person so I've been reluctant to pull the trigger yet.

So you hold these drivers in high regard?
 
I have not used them yet, in fact I only just unboxed one of the four to take a look.

Build quality seems very good. Judging from the specs, a pair of these is almost equal to a SB34NRX75 12" woofer, which I have used daily for two years. I think that 12 is excellent.
 
..High excursion is not needed and the enormous surround roll of CAC will excite IMD. Many midranges use inverted roll instead for a good reason...
Hi Juha -- So an inverted roll surround somehow doesn't excite the same level of IMD? That doesn't seem intuitive or logical, but perhaps you know something else? The Purifi piece doesn't make any comment about inverted vs outfacing (or convex?) roll surrounds.

Also, scanning mid drivers in Parts Express with an eye for inverted roll surrounds, sadly, I see only a few, contrary to your comment. (I personally prefer inverted roll surrounds if only for visual/cosmetic aspects. Any shower-cap style minimalist grill tends to touch the surround, which I think can't be good. And a grill that stays on the speaker is often an important aspect of WAF.)
 
Hi Jim,

Completely off topic, but did you know Captain Kirk has finally gone to space....

Back on topic, I follow your project with interest. Part of the reason is, I have the 15CAC too which I haven't put in a box. I was not aware of the Wondom solution, looks really interesting.

I am sure you probably already do that, but just in case, you might want to add in a capacitor in series with the tweeter to protect it from the sudden pop that may occur due to poor connections somewhere sometime.

Looking at your Sim, I would push the crossover to the tweeter a bit higher to maybe 4kHz, just keep it out of the telephone band. I suspect most mids crosses at 2kHz because of beaming factor and that many can't be pushed higher. But the 15CAC is something that can easily do that. Worth experimenting once everything is built since it is just programming... My experience with higher crossing is a more 3D sound..

All the best...

Oon
 
High excursion is not needed and the enormous surround roll of CAC will excite IMD. Many midranges use inverted roll instead for a good reason. Mids have better efficiency too, and highpass can be adjusted based on distortion tests in the box (easy again with DSP).
Read the surround excerpt more carefully.
Again the distortion is second order in nature and most obvious at low frequencies.
IMD is a symptom of nonlinearity, as is harmonic distortion. The main cause of the surround geometry variation with amplitude is driver displacement, primarily asymmetrical. This produces a second order nonlinearity, with associated second harmonic and second order intermodulation.
Most SB drivers have elevated bass distortion as measured, second harmonic mainly. I was under the impression that these SB drivers are a little subpar as far as the suspension is concerned, a possible source of that nonlinearity. In any case, Putzeys (and Risbo?) mentions Sd modulation, unlike the variation of Kms, can produce artifacts over the entire bandwidth of the driver. But as a midrange, there is much less required displacement to drive the surround variation in the first place. This is much like BL variation, which is driven by the displacement of the VC in the magnet gap and also produces artifacts over the entire range of the driver.
There are other reasons not to choose the 15NAC/NBAC/CAC as a mid, but Sd modulation should not be high on that list.

Hi Juha -- So an inverted roll surround somehow doesn't excite the same level of IMD?
The nonlinearity is reversed relative to the stroke, but not removed.
 
I have not used them yet, in fact I only just unboxed one of the four to take a look.

Build quality seems very good. Judging from the specs, a pair of these is almost equal to a SB34NRX75 12" woofer, which I have used daily for two years. I think that 12 is excellent.


Looks like we will both be building three way floor standers. I decided to get these as well. Four of these 8" woofers, one 6 1/2" TeXtreme midrange and one TeXtreme tweeter about rounds it out for me. I think they should play nice together. The TeXtreme drivers are excellent at the upper octaves but they do lack on the low end. I think the polypropylene cones will match the timber of the TeXtreme drivers a little bit better then the silver flute paper & wool cone does that is in my test box right now.

So where did you get the drivers from? Madisound? They only had one left when I enquired but I found four drivers at Meniscus Audio. They are on the east coast and I'm in California but I should have them by the 26th.

Good luck with your build. I will probably start a build thread too but I will wait until I'm almost done building them before I do because I'm working so much I hardly have any time to build these and it could end up being weeks in between posts if start a thread now.
 
That sounds like a really good combination. The MW16TX rather average when used as a bass driver, but as a midrange it is very excellent. I found I could cross it as low as 160 Hz before it started to show its low frequency limitations.

I found the MW16TX and TW29TX to be an almost magical combination. I experimented with crossover frequencies/slopes from 1.6k to 2.2k, and both 3rd order and 4th order. The differences were very small.

Yes I got the drivers from Madisound. I must have gotten the last batch.

Good luck with your project !
 
Mine where just delivered yesterday afternoon. They look like real nice drivers. Now I just need to modify my test box to fit the extra 8" driver since it was built with accommodations for only one 8" driver as that was what I was originally planning to use.

Have you done any measurements on any of the ones you have yet?
 
Btw Jim, during this process/planning, did you look into doing it with triple 8's?

For this project? No. I was specifically looking for a way to use the 4-channel Womdom JAB5 amp, so a TMWW tower seemed the obvious choice.

I have often wondered about those triple-8 speakers, such as Revel. I am guessing they are using 12 ohm or 16 ohm drivers, wired in parallel. Does anyone have insight into the woofers used in the Revel F328Be?

For all of us in DIY, we have don't have the option of custom wound voice coils, so we generally have to work with 8 ohm and 4 ohm drivers. A pair of 10's beats triple 8's most of the time, and as long as the cabinet is at least 11" wide anyway, the 10's will fit. But if there were good drivers available with 12 or 16 ohm voice coils, I might be tempted...

j.
 
But if there were good drivers available with 12 or 16 ohm voice coils, I might be tempted...

j.
There are lots of 16 ohm drivers made for the professional market but only BMS makes them in 8" and they are more mid oriented.

B&C does the 10CL51 in 16 ohm. If you step up to 12 or 15 inch drivers or go down to 5 or 6 inch drivers then there are lots.

They are made for where they think the pro market will want to array drivers.
 
Member
Joined 2005
Paid Member
Hi Jim,

As you are probably aware, the Textreme version of the 9.5” Satori woofer has been prototyped. WO24TX-4, WO24TX-8.

The SB23NBAC-8 (or slight variant) was used in the Revel Performa3 F208. In fact, SB Acoustics did their whole lineup of drivers for that whole Performa3 range.

Then Revel went off and got the drivers for the Performa BE Made in China.
SB were miffed, to say the least.

Mark Thomsen is the International Sales and Marketing Manager of SB Acoustics.

I will contact him and ask for a 16 ohm version.

And will design a 3 way with triple 8” woofers. The Satori is the leader but the standard SB23 is no slouch and probably the volume seller. I’d prefer shorting rings in the SB23, I will ask for that.