Active Compliance Management

It would be worth simulating with akabak, but i'm too lazy :D
Papasteack,

Thanks for bringing this technology to the forum! The dual 15 inch ACM™ frequency response of 32 – 169 Hz (+/- 3 dB) is particularly impressive for a bandpass design. Though the patent was filed in 2012, this is the first I'd read about it.

Although the output of a dual chamber bandpass box could be simulated if you had the parameters of the drivers, the asynchronus signal algorithms used by the ACM (Active Compliance Management) to individually drive each loudspeaker would also need to be included to get a valid simulation.

Art
 

Attachments

  • Asynchronus Signals.png
    Asynchronus Signals.png
    156.6 KB · Views: 260
Yes, but i feel it's simple delay/all-pass tweak. If we had vented Nd in multi entry tool in hornresp, it should be easily simed with delay and filters (Even if hornresp lack of all pass filters). Still actually multi entry tool could sim a vented enclosure with a bp4 firing in black chamber of the vented box. Should be yet fun to play with, with avantage of xmax limiting the bp4 driver compliance helper :D Do i need to patent it ? ^^
 
Last edited:
One great property of linear systems is that you can superposition their outputs. The opposite is also possible: divide the ACM thing into:


- a double chamber vented enclosure, with a shorted woofer in the divider
- a 6th order bandpass enclosure, with one port replaced by a port + a shorted woofer


Neither enclosure breaks Hoffman's law. Whatever way you drive them, driving them together will add +6 dB at most.
 

Attachments

  • Illistration-Typical-v-hs-221-02.jpg
    Illistration-Typical-v-hs-221-02.jpg
    34.9 KB · Views: 198
Last edited:
The thing i d like to see is direct radiator speaker excursion plot compared to without the second driver but with more juice to compensate in low end with eq. And spl contribution of the added port front loaded with second driver... I suspect this arrangement help getting shorted ports, so it push higher in frequency port resonnances. It's not hard at all to sim in older version of akabak, but compared to hornresp dynamic slidders tweaking ergonomy, it is really hard to get the set of parameters in the sweet spot.
 
I read the patent and could not find a place where they explain how they got 15dB free at the low end.
I thought patents were not granted if the workings and advantages were not explained and the invention should also be practically possible to implement.

So I went ahead to analyze with WinISD. I simmed 2 alignments using same driver to see xmax, phase and SPL.

Now, when two drivers share a space then the shared space can be reduced if the drivers produce different phases on the rear. If two drivers (magically) produce 180 degrees different output on the rear and same output in the front then the space behind them can be made zero. Of course, in case of normal vented boxes if we change the phase on the rear of one driver then the phase of that driver in the front too changes wrt the other driver, some thing we dont need. The fronts must track in terms of phase, thats the goal. We only need difference in phase at the rear.

Now, if we add an acoustic low pass on front of the drivers then the phase of the front radiation from that driver changes due to the acoustic low pass but the phase of rear does not, this is the key. Lets see the phases of a vented and a 6th order bandpass. A vented box has +180 degrees at tuning, say 35Hz and zero phase (close to zero) higher up. Compare this to a bandpass, which has +180 degrees at lower tuning (same as vented) and ~-180 degrees on the upper side. Now thats interesting, this yields the following

At 35 Hz: vented is at +180 and bandpass too is at +180
Between 35 and 50Hz: vented is moving to zero and bandpass is moving to -180, changing faster than the vented.
Above 50Hz:Vented is at zero (almost ) and 6th order bandpass is -180.

Note the last combination, the rear outputs are in phase but the fronts are not, they are going to 180 degrees difference as we go up the frequency. Now in order to equalize the acoustic phases we could delay the vented depending on the frequency using a DSP, lets say starting from 45Hz and moving higher up in frequency. This will equalize the acoustic output and create the same difference in phases on the rear that we had in front earlier. A phase difference at rear means a small box so, yes, box reduction above 50Hz is possible. I have not attempted to analyze, how much the reduction is. Also, is that really an advantage? May be minor gain in size.

There is another dimension to this. When we look at constant SPL from both the alignments, we see that the excursion is not constant at same frequency, in fact, the excursion curves of both the drivers are same only at 35Hz and totally different everywhere else because they are in different alignments. Maybe this could be used to maximize SPL, ie use driver that produces more SPL for the given xmax at that frequency thereby generating maximum total output. While I dont have curves to post, I did analyze this using WinISD but did not get anywhere near the 15db claimed.

Either, I just didnt get it or the patent is questionable.
 
Last edited:
Translated from "Marketing Speak" to "English" I think this may be what's going on here:

2-fig-1-HS-28-Subwoofer.jpg


1) both woofers share a common chamber

2) If you juggle the variables carefully, you can configure the two woofers so that one is 90 degrees out of phase with the other. If you do that, it means that when one cone is fully compressing the air in the chamber, the cone of the other woofer isn't compressing the air in the chamber at all. One could argue that this is equivalent to "increasing the chamber size by 200%" since neither cone is every compressing the air in the chamber fully. When woofer #1 is compressing, woofer #2 isn't, when woofer #2 is compressing, woofer #1 isn't, etc.

I don't think this could be modeled with Hornresp or WinISD, but Akabak can probably do it, even the old version.

Their wild claims of a 15dB increase in output are obviously silly; the only way they can get to that number is if they truly treat it as if the box volume is doubled and they're looking at a single frequency below tuning frequency. IE, ported boxes have a rolloff of about 24dB per octave, so if you doubled the size of a ported box, you could potentially get a 15dB bump in output IF it was already tuned too low.

If you look at the spec sheet, the output levels they're getting aren't much different that what competitors are getting with their boxes of comparable size, so I don't think they've broken Hoffman's Iron Law.

This isn't meant to "trash" the design. Possibly the biggest innovation there is that you could probably get away with lighter weight/flimsier construction if the chamber is never pressurized 100%. This was an issue when I made a tapped horn back in the day that was tuned to 15Hz; the wavelengths are so large that even with just 150watts per woofer, the cabinet was visibly flexing. A combination of cheap construction on my part, along with me failing to realize that a 15Hz resonance really flexes the heck out of a cabinet, especially when the cabinet itself is resonant. If you put your hand on spots in the enclosure, you could actually FEEL where the pressure maximums were located in it.
I don't think that the VUE subwoofer would require particularly complex DSP. You want the two woofers to be 90 degrees out-of-phase. To me, it looks like they accomplished half of that, by the use of the bandpass chamber. Basically the output of the bandpass chamber has to travel 45cm further, because it's in that chamber. The second half of the delay could be done with DSP, about 45 degrees of delay. The delay would only be optimized for a single frequency, unless they're using FIR filters.

Once could accomplish something similar in a tapped horn by seperating the two woofers by 1/8th of a wavelength at the lower tuning, and then adding a DSP delay of another 1/8th of a wavelength. I can't recall if Hornresp offers the ability to do delays. If you were really masochistic you could even do it with plain ol' crossover filters. Just use an asymmetrical low pass. Basically stagger the low pass filters so that one woofer has more delay than the other.

EDIT: LOL at the fact I simmed this three years ago already. Whoops.
 
If you do that, it means that when one cone is fully compressing the air in the chamber, the cone of the other woofer isn't compressing the air in the chamber at all. One could argue that this is equivalent to "increasing the chamber size by 200%" since neither cone is every compressing the air in the chamber fully. When woofer #1 is compressing, woofer #2 isn't, when woofer #2 is compressing, woofer #1 isn't, etc.
That's looking at the interior of the box. But what about the sum of the output side of the two drivers?

A Karlson has two (or more) chambers in series and other fixings. Despite the "dog's breakfast" design, it has remarkably nice sound (amazing for cellos), which is why it seems to have continuing enthusiasts.