What killed off the acoustic-suspension speaker?

[alspe]

Question for anyone: I'm not exactly sure how an aperiodic design (like the Dynaco A-25) would be classified, but would the modern version* of its woofer SEAS sells now be a good candidate for an acoustic suspension enclosure?

Seas Prestige A26RE4 (H1411) 10" Paper Cone ($138 bucks....yikes)

The driver's rubber surround looks like something a subwoofer driver would use! And cone excursion - 8mm peak to peak - seems good.


* SEAS says this newest version of the new version will no longer fit in an A-25 because the frame is now larger by 1cm - I wonder why they did that?

Maybe it is quite for a.s.

OFFTOPIC, Seas A26RE4 has quite large inductance (3,4 mH) (about 3-4 times as much as other drivers). So it includes low pass filter (iron core inductor) for use it in 2-way system. If I have understood correctly, inductance should always be low in good drivers? Is large inductance driver hard load for amplifier ? Inductive load at bass region?

 

[Esperado]

If I have understood correctly, inductance should always be low in good drivers? Is large inductance driver hard load for amplifier ? Inductive load at bass region?

The only problem with inductance of loudspeaker is the saturation of the magnet/core at hight levels. Better to look at the response curve and distortions.
It is always a good practice to Zobel and compensate motional impedance. Specially in closed enclosures where resonance peak is huge.
As long the enclosure's impedance is flat, your amplifier will not have to deal with stranges V/A, damping will be better, phase too and internal serial resistance of the amplifier+cable's resistances will not modify response curve.
That is one of the good practices, when it is about to build a loudspeaker's system, along with avoiding like hell closed enclosures ;-)
I do not understand why some people, here, believe that BR makes some sort of "boomy" sound. It is not true at all. If you could ear how my BR system is "fast" dynamic, dry, airy, with no resonances at all, no one can confuse between kick drum and basses notes....
Who want to loose near one octave at the low end, and kill life in medium ?
Make a simulation with this 10" Beyma speaker in quarter waves line and the same in closed box, who can hesitate, even half a second ?

To answer the question of the OP:What killed off the acoustic-suspension speaker?
my response would be:
"Because it is stupid ?"

 

[alspe]

From another thread :

Your assumptions about a speaker for a closed box design are not all correct. Choosing a lower sensitivity driver will be a good trade-off for better depth, but not when the driver's Qts is rather high.

Often this situation occurs but what you need for a good closed box design is a woofer with strong drive and low Qts, together with a soft suspension and very low Fs. Then put it in a box with a volume of about half or less than Vas to gain a Qtc of 0.5 - 0.7. In this situation, the compliance of the box dominates over driver compliance, better reaching the theoretical maximum out of a given box.

Putting a high Qts driver in a box > Vas is definitely not the way to go...

What parameters shows that driver's suspension is soft? Can that kind of driver play even if it has low Qts?

 

[Esperado]

What parameters shows that driver's suspension is soft?

Light Moving mass and low FS
Of course very low FS is the first thing to look-at, because closed box will increase-it.
I had gave-you a link to a french web site where you can simulate in line all the enclosures, including ports sizes and box designs from a given loudspeaker. They have a huge database of Thiel and Small parameters of a lot of loudspeakers.

But don't stay stuck with closed enclosures. The one and only real advantage their offer is the 6 db/oct under their cutoff instead of 12db for B.R. So you can go deeper with amp compensation, that you will pay with an important coil elongation (X).
Remember, in a closed enclosure, you lose half of the energy produced by the loudspeaker. and a (very) little dynamic in the medium because the added charge of the air's compression.

Remember too, if you are not about planing to play organ pipe at 16 hz, that the lowest bass note is 41hz. No need to go further. No need too to introduce distortion in moving the cone at lower frequencies, and there 12db/oct is better again. I don't see any low and huge movement on my loudspeakers listening at normal (high) levels.

I think you are on the wrong way to look at a loudspeaker for a given (closed) enclosure. Better to find a loudspeaker witch suit your needs (response curve, distortion, max SPL, power handling, efficiency) and find the best enclosure for him.

 

[alspe]

Light Moving mass and low FS
Of course very low FS is the first thing to look-at, because closed box will increase-it.
I had gave-you a link to a french web site where you can simulate in line all the enclosures, including ports sizes and box designs from a given loudspeaker. They have a huge database of Thiel and Small parameters of a lot of loudspeakers.

But don't stay stuck with closed enclosures. The one and only real advantage their offer is the 6 db/oct under their cutoff instead of 12db for B.R. So you can go deeper with amp compensation, that you will pay with an important coil elongation (X).
Remember, in a closed enclosure, you lose half of the energy produced by the loudspeaker. and a (very) little dynamic in the medium because the added charge of the air's compression.

Remember too, if you are not about planing to play organ pipe at 16 hz, that the lowest bass note is 41hz. No need to go further. No need too to introduce distortion in moving the cone at lower frequencies, and there 12db/oct is better again. I don't see any low and huge movement on my loudspeakers listening at normal (high) levels.

So suspension compliance mm/N should be quite high? Many mm's with low force (N)?
I have WinISD so I can sim drivers. To me, it looks like that highish Qts drivers go low but they require enclosure volume...

If I undestand correctly, rear radiation of driver is 180 degrees wrong phased so reflex gives you more bass energy that is "faulty". If that is heard is a another matter but in theory it seems to be wrong. True it can be that too low closed box eats dynamic because of air compression.

Low tuned bass guitar E is 31 Hz. If you tuned your medium sized reflex speaker around 40 Hz there's almost one octave below it where driver can be damaged because of huge movement. Closed has always some kind of limiter by air spring. True is that cone moves less around one octave wide area between tuning in reflex.

OFTOPIC : But there is complexity of reflex port, designing it so it is big enough, long enough, fits in enclosure, does not make resonance sound in midrange and that air can move in it (damping inside enclosure). What I have look at measurement of speakers is that if smallish speaker is tuned semi low with port, its output seems liked closed box with low Qts driver. There might be a lot of faults that air "finds no way to port" because small enclosure is too much stuffed, port opening is too close to wall etc.

 

[alspe]

Must add a remark. What if you want to build a Closed Box having a decent deep bass extension like -3dB at 45 Hz, and a decent power capability in the bass ?

You need a long-throw 12 inch woofer with rubber periphery. It is quite expensive. If you want to keep the overall cost reasonable, you need to stay two-way, which means a woofer and a tweeter. And that's terrible. Because with a 12 inch woofer, the distance between the woofer and the tweeter emission centres is too big. You'll get a bad stereo imaging and you'll endure fatigue while listening. And more terrible, a 12 inch woofer stops working correctly above 2 kHz. You thus need a steep crossover (say 4th-order), somewhat expensive, operating at something like 1400 Hz. Which causes the next issue : the tweeter needs to be quite large (say a one inch diameter tweeter, these are not the cheapest) for coping with low crossover frequency.

For the same amount of money you can design a Bass Reflex using two 6 inch woofers in a d'Appolito configuration with an inexpensive 3/4 inch tweeter crossing at 4 kHz, possibly 2nd-order. It will deliver 45 Hz at -3 dB. It will handle the same power in the bass like the above Closed Box design. It will behave much better above 2 kHz. And being narrower (6 inch woofers instead of a 12 inch woofer), it will integrate easier in your interior.

You can not compare 2x6 inchers to one 12 incher. 12 incher in wider box gives way better upper bass slam, directivity and has almost double cone area. And moves air better than two separate small drivers. 2 drivers parallel makes ~ 3ohm bass load which is not the easiest for amp.

10 inch bass with waveguide tweeter might option.
Aurum Cantus AC-250F1 does not look like the worst driver. Goes nicely in 40 liters which is not bad for 10 incher. F10 31 Hz, F3 around 50 Hz though...

 

[steph_tsf]

... better upper bass slam & directivity ...

Can you explain ? Do you mean a more accurate transient response maybe ? Directivity in the bass range ? Are you talking about the huge subjective difference between a closed box and other solutions like a 8th order bandpass aka "Bose" ? I may agree ... that's a very interesting theme.

 

[alspe]

^ wider enclosure - lower baffle step frequency, more lower midrange energy towards listener.

 

[DBMandrake]

So suspension compliance mm/N should be quite high? Many mm's with low force (N)?

Yes

I have WinISD so I can sim drivers. To me, it looks like that highish Qts drivers go low but they require enclosure volume...

Qts is not directly related to cut-off frequency. For an acoustic suspension design the Qts will rise a long way from its free air value, so you want a free air Qts quite a bit lower than the desired box Qtc.

You're looking for a driver with high compliance, heavy cone, large Vas, low Fs.

Since you have WinISD, just try siming a lot of different drivers and see how the responses you can achieve follow on from their parameters - you'll start to get a feel for what you're looking for.

Low tuned bass guitar E is 31 Hz. If you tuned your medium sized reflex speaker around 40 Hz there's almost one octave below it where driver can be damaged because of huge movement.

Closed has always some kind of limiter by air spring. True is that cone moves less around one octave wide area between tuning in reflex.

31Hz is not one octave below 40Hz An octave is doubling or halving frequency, so an octave below 40Hz is 20Hz.

Your point is correct though - a reflex box tuned to 40Hz will start to lose loading below about 35Hz.

 

[bocama]

From another thread :




What parameters shows that driver's suspension is soft?

That is the VAS it is the equivalent volume that has the same suspension as the driver

 

[Esperado]

That is the VAS it is the equivalent volume that has the same suspension as the driver

YEEEEES ;-)

So suspension compliance mm/N should be quite high? Many mm's with low force (N)? .
I have WinISD so I can sim drivers. To me, it looks like that highish Qts drivers go low but they require enclosure volume...

Yes and yes.

If I undestand correctly, rear radiation of driver is 180 degrees wrong phased so reflex gives you more bass energy that is "faulty". If that is heard is a another matter but in theory it seems to be wrong. True it can be that too low closed box eats dynamic because of air compression..

You kill-me. It will be a long run talking to phase and group delay, and will be not about degrees in one phase, but how many phases ;-) A matter of distance, with speed propagation of sound in the air plus waves lengths, in nature, and a matter of Qt in enclosures. An extremely complex thing to discuss, as even filters, loudspeaker's characteristics, phase response in amplifiers, vertical alignment between transducers introduce phase rotation and, so, changes in group delays. What can i say to resume ? A B.R. where Fr= Fs is the best for transients, and, yes, closed boxes, in certain conditions can be better (or worse). It is all about what i called so stupidly "fast basses".

Low tuned bass guitar E is 31 Hz. If you tuned your medium sized reflex speaker around 40 Hz there's almost one octave below it where driver can be damaged because of huge movement. Closed has always some kind of limiter by air spring. True is that cone moves less around one octave wide area between tuning in reflex.

You are right, but we have to remember than x is directly correlated with response curve and level. Here, too, filtering is the answer: reason why i'm on the way to port my system under a DCX2496 numeric filtering and active amplification, where i can play with filters and delays easier.

OFTOPIC : But there is complexity of reflex port, designing it so it is big enough, long enough, fits in enclosure, does not make resonance sound in midrange and that air can move in it (damping inside enclosure). What I have look at measurement of speakers is that if smallish speaker is tuned semi low with port, its output seems liked closed box with low Qts driver. There might be a lot of faults that air "finds no way to port" because small enclosure is too much stuffed, port opening is too close to wall etc.

Right again. Plus noise of the air moving in the port because speed, depending of the level... But rare are the occasions where you cannot find a solution with more than one port if necessary.
At the end, building an enclosure is such a lot of compromises, you have to deal with so much parameters, that, at the end, it is a matter of good habits+feeling.

Ps: I just made a group delay simulation with the same loudspeaker in the same volume enclosures. It gives:
Sealed: 6 milli-seconds @ 38Hz
Bass reflex:10 milli-seconds @ 25Hz
Bandpass: 14 milli-seconds @ 50Hz
Some agree that, under 20ms, one cannot ear any differences. Not to worry too much about.

And, here, what a friend of mine had obtained with a DCX and an 3 way bass reflex enclosure:
http://www.dcx2496.fr/img/Group_delay.jpg

 

[alspe]

31Hz is not one octave below 40Hz An octave is doubling or halving frequency, so an octave below 40Hz is 20Hz.

Your point is correct though - a reflex box tuned to 40Hz will start to lose loading below about 35Hz.

Yes, I know. (I don't usually write english so my output can be bad)
I meant that if you tune at about 40 Hz (easily can not go lower in small-mid size boxes) there is an octave where port do not protect the driver.

YEEEEES ;-)
Yes and yes.

You are right, but we have to remember than x is directly correlated with response curve and level. Here, too, filtering is the answer: reason why i'm on the way to port my system under a DCX2496 numeric filtering and active amplification, where i can play with filters and delays easier.

All filters make phase changes etc if I have understood correctly? Subsonics, bass refles is high pass filter? or is it?
DCX2496's aren't for me. I'm (always) looking for and simple setup cd-amp-speakers no knobs to tune anything, just eject and play button on my cd-player
Heard so many stories that they (DCX...) are blurring sound.

 

[Esperado]

All filters make phase changes etc if I have understood correctly? Subsonics, bass refles is high pass filter? or is it?

All analog filters turn phases to be precise. Loudspeakers give a progressive phase switch up after its resonance frequencies because the inertia of the mobile mass.
Not a problem as long as it is progressive and linear with frequency, equivalent to adding a distance between source and listen position. Nothing to worry (and nothing we can easily modify) in the low range. One good way to approximate the vertical alignment of speaker is to align vertically the moving coils. You can tune-it more precisely in looking a square wave at each cut-over frequency. Here my enclosures:

Most of the market's baffles are not time aligned at all, with all transducers on the same vertical plan, whatever the depth of their cone and some are sounding in a agreeable way, proof that is not so important than that. But, when you align, you get better transients and sound coherence, for sure.

DCX2496's aren't for me.
Heard so many stories that they (DCX...) are blurring sound.

They are cheap, and some poor parts (PSU, analog, clocks) need to be tweaked. Once done, will be better than most hight end CD DAC and the filters are in the numeric world 24 bits/96 kHz. You can equalize your system to +- 1db, with the kind of dream group delay you have seen.
Their offer 3ways crossovers , each slope of 6/12/18/24/36/48 db/oct following different curves, phase slopes, and 31 parametric equalizers by way, global delays on each way to compensate vertical alignments, and an automatic program to tune-it (in the time domain). You can record all your programs, compensate room's resonances, etc...and get incredible improvements, reduce volume of your boxes, it is just a dream box.
And all that for near the price of good air selfs and condensers you would put in 24db/oct symmetrical crossovers ?
Now, with the incredible low prices of so good D amps Kits, there is no reason, even financial, to hesitate, and people telling fairy tails about DCX are just ...telling... fairy tails. The same people will buy some very fantastic speaker cable, adding transparency and dynamic because liquid helium cooled and silver plated (mouaaaarf) for 10 times it price ?

 

[Volenti]

Yet again rather misleading. "any arbitrary response" ? Really ? So you could get a box Qts lower than the drivers free air Qts ? No, I didn't think so.

There is nothing magical about using fibrous damping - it's just one source of damping within the box at bass frequencies, and not the dominant one. Because of the way fibres affect the adiabatic process of air compression and rareification, it's possible to make the box "look" slightly larger than it really is - up to about 15%, but there is no sonic benefit to this, it's just a different way of achieving the same response that requires a slightly smaller box. (Improvements in midrange standing wave absorption can be dramatic though)

Actually yes, you can get a lower in box QTS with damping material than the driver has in free air, granted this takes a driver with fairly high QTS to begin with (0.8+ usually) but it is possible.

And with your second paragraph, you're completely ignoring the resistive damping the fill does, I'd call halving the impedance peak at resonance (compared to a bare enclosure) a pretty dominant damping source.

 

[DBMandrake]

Actually yes, you can get a lower in box QTS with damping material than the driver has in free air, granted this takes a driver with fairly high QTS to begin with (0.8+ usually) but it is possible.

If you want to be picky, with a really large closed box with a lot of stuffing and a driver with an unusually high Qes relative to Qms you might see a slight decrease in overall Qts compared to free air.

However acoustic suspension designs were specifically being discussed not just any arbitrary closed box, which means ~ Vas/Vb >3. Show me any acoustic suspension design with that tuning ratio which would allow the small realisable reduction in Qts from stuffing to override the much larger increase in Qts from the extra stiffness added by the box.

And with your second paragraph, you're completely ignoring the resistive damping the fill does, I'd call halving the impedance peak at resonance (compared to a bare enclosure) a pretty dominant damping source.

How am I completely ignoring the resistive damping of the fill ? I did say it was one source of damping in the box at bass frequencies did I not ? Do you know of another kind of damping than resistive ? By definition damping is resistive, whether electrical resistance or mechanical.

I also stand by my statement that added fill in a closed box is not a dominant damping factor at bass frequencies, only a supplementary one. In most decent drivers Qts is dominated heavily by Qes when driven by a voltage source.

Seeing the impedance peak halve when fill is added is all very well, and shows that there was a significant reduction in Qms, but if Qms was many times higher than Qes to begin with (which is usually the case) then the overall Qts value will drop very little. The impedance curve might change a lot but the frequency response will change very little.

Increasing mechanical damping significantly doesn't make much difference if electrical damping was already dominating, which would be the case unless you had a really high Qes (cheap driver) or were driving it from a high impedance source...(valve amplifier)

 

[Esperado]

I wonder if any complete scientific studies had been done on different damping materials, and positions of it. We just have the habit to approximate to near 10% the relative augmentation of the volume a full fill gives in a closed box, with glass wool, or think it will compensate near the HP's volume. In a B.R., positioning the damping material change drastically the low medium response curve. And that is not taken in consideration in any calculation tool.
At the end, building enclosures stay very empirical, need measurements in the real world to tune-it ... So many parameters with mutual influences...
I wonder why, with so poor measurements (distortion, phase etc.), comparing to electronics's performances, our old loudspeakers are able to sound so good, sometimes...

 

[River757]

The OP's question has been answered a bunch of ways, but I recall something about this in the Dickason book.

Yes it has - thanks to everyone who contributed!

Speaking of books, I was one of the many amateur speaker builders who got their start by reading the David Weem's publication "Designing, Building, and Testing Your Own Speaker System, with Projects" which is still in print. Fortunately for me it did not delve too heavily into the mathematics of speaker design but still managed to teach me a fair amount of basic audio theory.

 

[Esperado]

The OP's question has been answered a bunch of ways.

Yes it has.

Colonel Mustard in the kitchen with a rope ?

 

[River757]

Colonel Mustard in the kitchen with a rope ?

Clue!

Btw I was looking for pics of vintage audio and ran across this cool photo of the AR9, Power Sponge Of The Decade , which contained a link to the designer's thoughts on the engineering behind the 9.

Some of Infinity's a.s. speakers were also known to be difficult loads for many amplifiers, including the beautiful RS IIA, which IIRC also exhibited tricky phase angles electrically-speaking, though to be honest I don't know what that means or how it affects an amplifier.

 

[Vintage Pioneer Guy]

All I know is that I have yet to find any speaker sounding better than my "sealed" BOZAKS period,custom built or otherwise! Concert Grands,4000 Symphonys or some of the smaller series for me can't be beat