[JeffB]

I was reading about the AV123 LS9s here:
http://av123.com/index.php?page=shop.product_details&flypage=shop.flypage&product_id=185&category_id=36&manufacturer_id=0&option=com_virtuemart&Itemid=37
Specifically the section titled
"Speaking of Drivers... let's talk about bass for a moment..."

AV123 lays out the argument that its 12 6.5" woofers will sound better than any 12" speaker.

So I am curious about the general question.  Can better base be obtained from multiple smaller speakers than a single bigger speaker?
Is the answer to this question any different for Open Baffle designs?
I am asking as curious layman.

I am not so sure I believe the argument layed out by AV123.
I have a 15" woofer on an open baffle and from 100Hz down it is some of the best bass that I have heard.
It is better than my sealed sub box with two 10s.  It is better than a ported speaker that I have with two 8s.
I believe it to be better than the Paradigm Studio 100s I demoed with four 7s.
It is better than all the Polk speakers I demoed at Fry's.

I have never done a side by side comparison or double blind test, so I could be mistaken.  This is just from memory.

AV123 states that a bigger woofer will have more mass, so is harder to stop, and will not as accurately handle note decay.
In the case of the 15", I can definitely imagine this affect above 100Hz, but I am not convinced it is relevant below 100Hz.

I am also puzzled how multipe 6.5" drivers even make bass.
I understand that there is lots of surface area to move air.
However, the response of most single 6.5" drivers drops off fast starting around 150Hz to 200Hz or so.
I have a hard time envisioning the responses down around 30Hz adding up.
Would such an arrangement need a contour filter to pull up the bass?

In my case, I have an 8" Visaton B200 mated with a 15" Warrior on an 18" baffle.
I am essentially lacking that hard hitting base thump of the kick drum.
I get sufficient SPL in the 100Hz to 200Hz region, but I don't think I have sufficient control in this region.

So I am pondering whether to try a different 15" driver like an Eminence Alpha or to go with two 12s or four 10s.
The Warrior 10s seem like a nice bargain.

[JeffB]

Just after posting this, I went looking for info on comb filtering.
I found this
http://www.roger-russell.com/columns/columns.htm#single
Twenty 3.5" drivers.
Response 20Hz to 18kHz with EQ.

[JeffB]

Four Warrior 10s is starting to tempt me.
$30 each.  So $240 total for left and right speakers.
I assume I can wire them such that pair A is in parallel and pair B is in parallel and
A and B are in series.

This is twice the surface area of a 15".

[Graham Maynard]

It does not matter which driver(s) you use, they will roll off below resonance, and exhibit phase change about that resonance.

This is why I tried a circuit which can boost below resonance.
http://www.audiocircle.com/index.php?topic=53322.0

It also changes waveform phase which makes the kickdrum sound much more like it should, which a single circuit LF driver(s) on a baffle cannot do.

You've already got a 15", probably a decent driving amp.  Do you have a transformer and capacitor from a PSU, and say the LF choke from the other channel you can try ?

Cheers ....... Graham.

[markC]

Try driving those 15"ers with a seperate plate amp. There's no lack of kick drum response for me. When I get in the mood to play a little loud, the wife gets pissy about the kick drums.

[Graham Maynard]

Hi Jeff,

I've run through some drivers using Xlbaffle;-

http://server6.theimagehosting.com/image.php?img=001.129.jpg

http://server6.theimagehosting.com/image.php?img=007.662.jpg

http://server6.theimagehosting.com/image.php?img=008.918.jpg

(click on an image to clarify text)
(one of the Fane traces was overwritten, but driver not noted)
Any driver which uses 'Qes' to boost low frequency output tends to miss out on an initial 'hit' when first getting the cone moving, thus any driver which exhibits even the slightest bump in the frequency response might sound a little lacking.

My favourite from these curves is the Beta-15A, but it is easily excursion limited at LF.
The Warrior 15" has a slight bump, though much better power capabilities.

Does anyone have another suggested driver for LF drive as well as LF extension worth examining which will not make a big hole in the pocket ?


Cheers ......... Graham.

[JeffB]

Thanks Graham.

Are these measurements for an infinite baffle?
When you say 4x is that two on each baffle or four one baffle?
The Fane 4x12-125 is graphed in both the first and second link with very different SPL measurements, so why the difference?

I have tried a few times on this forum to get some useful understanding out of these SPL charts, but nobody has yet really been able to help me out.
My basic assumption is that flat is best, and since we are talking about bass drivers, I would think flat below 100Hz is what one would look for.

The Fane 4x12-125 looks flat to almost 40Hz in the first link.
The 2x15 Warrior looks flat to about 45Hz.
So by my reasoning these look about the best.
However, you like the 2x Beta 15A, which starts dropping at about 100Hz.
It is only down about 4db at 50Hz and 6db at 40Hz, but visually it looks like a much faster fall-off than say the Warrior.

So what makes the Beta 15A look better to you?

[JeffB]

Hi Graham,

I would like to see a plot of the Warrior 10". $30.
Specs here:
http://www.shredmuzic.com/product_p/813-019.htm

DMason brought up this driver yesterday AE IB15
http://www.aespeakers.com/drivers.php?driver_id=8

It looks interesting for $100.  I can't find all the specs though.

[ttan98]

AE IB15 from AE has low SPL, 86dB other parameters seem ok.

[Dmason]

I feel that two IB15 + a high power switching plate amp will produce bass, impact, dynamics, and all those other things. It gets my vote, and order, incidently. That thing will move ALOT of air. Frees one to use low power tube amps on the mains driver.

Graham, -I used 4X Delta 15LF on all DarkStars, 20 cones in all, and they work very well. The Beta does look better on paper.

[Graham Maynard]

Hi Jeff, ttan and Dan,

I will see what more comparisons I can run off, though Xlbaffle is easy to download, and it shows how a corner/room peak can develop around 100Hz which is not directly related to the driver.

Jeff - your's is a good question about what Xlbaffle really shows, for it can be steady-sine based only, and not directly relate to the dynamic responses necessary for Dan's 'fine progressive rock'.
Neither the Alpha nor Beta 15As do as well as the Gamma-15A in MJK's impulse response tests at;-
http://www.quarter-wave.com/OBs/OB_Design.pdf
where simulated baffle responses are also steady-sine based only.

I think the correct 4x 12-125 Fane plot is the near flat one !!!
When I first saw this I thought 'Wow', but then I realised that this is too flat for high impact dynamic bass because that flatness before roll-off is due to electromechincally uncontrolled resonance which cannot fail to reduce dynamic impact because that same lack of electromechanical control confers a weakness in leading edge attack.

I'll check/compare those drivers just mentioned.

Jeff. 
Erling tried my transformer circuit;-
http://www.audiocircle.com/index.php?topic=53322.msg480890#msg480890

so maybe your Warrior could do better ?

The thing is, you might buy 4x 10" or another 15" and still find your room corner still gives a similar unpleasant peak near 100Hz which my circuit takes out.
Again though, any modification of the frequency response simultaneously modifies the dynamic response, so a working compromise becomes necessary, which is possible via component value adjustment.
To my ears the transformer circuit improves reproduction through reducing electromechanical driver storage due to series impedance increase and thus reduced drive at the resonant frequency, whilst boosting drive below that.

Hi Dan,

I see the Delta-15LF is not expensive, has longer throw for bass, cone not too heavy, impedance okay for LF only, and not a rubber surround - which always impart a dull rubbery sound to my ears !
How are you not a trembling wreck from all those cones ? 
You must be displacing 'gallons' of air !!!
Bet playback of a diesel loco or V8 stocker is mighty realistic.
http://freesound.iua.upf.edu/samplesViewSingle.php?id=35497
http://freesound.iua.upf.edu/samplesViewSingle.php?id=26831


Cheers ......... Graham

[Graham Maynard]

? PUZZLE ?

Now I must ask what it is that Eminence are showing on their driver SPL/W/m plots because it is as if their lowest real measurement point is actually at 170Hz and then they merely draw the same roll-off curve and 12dB/oct line as if this is the driver plot - which is absolute rubbish because this CANNOT take account of driver resonance and VAS characteristics.

In the attached simulations I have taken Eminence's 170Hz SPL figure as the reference point for Xlbaffle calculation which DOES take account of different driver resonance characteristics.

If I compare the simulation with Eminence's *assumptions*, then
relative figures for output at 30Hz are;-

Beta-15A           Em-75dB  XlB-85dB;     VAS 335, Qes 0.63
Kappalite15LF    Em-78dB  XlB-84dB;     VAS 159, Qes 0.41
Definimax15LF    Em-80dB  XlB-82dB;     VAS 115, Qes 0.54
Kappa-15A        Em-77.5dB  XlB84.5dB;  VAS 321, Qes 0.33

http://server6.theimagehosting.com/image.php?img=011.09d.jpg

Cheers ......... Graham

[Graham Maynard]

PUZZLE:  Part 2.

Looking more closely at the Eminence SPL plots I now conclude that the lowest frequency at which they publish real-world measurements is 200Hz.
Clearly any reflection or resonance ripples on a frequency response below 200Hz will be as much influenced by driver usage as by the driver itself, but I still think it wrong that Eminence draw a 'representative' line below 200Hz which simply cannot represent the driver's own natural characteristics.

I have attached a another set of four XlBaffle simulations, these based upon Eminence's own 200Hz SPL figure for each driver.

The Kappalite has a heavier cone and Neodi magnet, but also just over twice the X.max displacement of the others, so can displace twice the air!  However, two of the others in parallel will not only shift an equal amount of air as the Kappalite, but also move their SPL trace upwards by 6dB due to the cone area and amplifier loading both being doubled.

Where before I favoured the Beta-15A over an Alpha-15A, the bulge in the Beta characteristic shows that this driver generates increased bass output due to a 'natural' resonance which cannot fail to introduce its own 'voice' upon reproduction. 
A characteristic which limits its impulse capabilities too, as MJK revealed.

Multiple smaller drivers might not generate an Alpha or Beta like 'voice' and thus have a much better impulse response, but then the three other drivers shown here do not have that 'loose' bulge in the bass response.  Thus I cannot see any advantage in using several small drivers over the likes of the other three shown here, though I have still not looked at other makes.

http://server6.theimagehosting.com/image.php?img=013.458.jpg
Click on image to expand.


Cheers ............. Graham.

[JeffB]

Thanks Maynard,

I think I may finally be learning something, but I am not sure.

I am assuming that you think the Beta 15 has a resonance simply because its graph does not conform to the shape of the others.
That is these 15" drivers with the xmax that they have should roll off smoothly.
In order to extend the flat bass response either an 18" driver or a driver with a lot more xmax would be required.

[iON]

Howdy,

Regarding the original question:

I haven't heard the AV123's but obviously they are capable of moveing a lot of air. The effective area is larger than three 12-inchers. However, area is not everything. (Dont get me wrong though - I definitly have a weak spot for big transducers).

Without going into t/s paramater details, One immediate advantage I can think of is that the load is spread over 12 coils which is indeed beneficial in terms of power compression. Low power compression is always welcome in the quest for dynamics.

As for OB setup my first choice would not be 6,5 inchers though,.. but perhaps an array of 10-inchers..  

Cheers,
iON

[Graham Maynard]

Hi Jeff,

Musing more about LF drivers and particularly with regard to your already stated observations about the Warrier:

LF driver types tend to have heavier cones in order to compress cabinet air-springs without disintegrating.

On an OB the LF air loading is lighter so the heavier cones will absorb relatively more of the electrical waveform energy just to make them move and return, than will a lighter cone.

Cone movement and mechanical resonance are controlled by the voice coil, which in turn cannot perform better than its Qes figure (voice coil size and magnetic field related) no matter what the cone and air loading are.

I have done another XlBaffle study of sine amplitude based SPL for the Beta-15A with three other plots based upon the same driver as if its Qes had been modified.

Each of those plots being centred on 100Hz which is where we are likely to level the large to mid or widerange crossover points.

http://server6.theimagehosting.com/image.php?img=015.25e.jpg
(click on image to expand)

I'm sure we all know about critical damping and Q, which requirements the 0.83 trace here almost meets, hence the illustrated beginning of an amplitude peak before a steeper LF roll-off.

However we are not normally informed about energy storage and phase changes in relation to these oft presented amplitude plots.

A driver having a Qes as per the red trace here, is without doubt going to provide the flattest low frequency response, and would appear to be a best choice for baffle fitment, but, its 3dB of lift above that from a Qes = 0.43 driver arises due to cone mass and suspension spring storage during the first half cycle of energisation.  This potential energy is then released kinetically into the second half cycle, and becomes an on-going exchange before the steady sine energised response can be measured !!!  Hence too the normally illustrated driver impedance characteristic peaking due to stored energy at the resonant frequency, which needs less sine drive to maintain amplitude at that frequency once energised.  (Voltage drive, often 2.83V for 8 ohm drivers, is assumed.)

That potential energy which becomes stored by the driver during the first half cycle is thus denied as kinetic energy to cone movement and air pressure generation, so the impulse response is *degraded* as the amplitude response becomes 'improved'.  As shown, but not explained, by MJK on page 10 of;-
http://www.quarter-wave.com/OBs/OB_Design.pdf

Also, when that stored energy is returned by the higher Qes driver it is *always* within its natural Fs based frequency range, no matter at what frequency of drive the displacement energy was stored in the first place !!!

This is why the higher Q bass drivers impart 'tone', and why you cannot enjoy clean hard rock through them because the loss of dynamic attack and the narrow bandwidth of returned energies simply muddy the whole sound.  Higher Q drivers can be adequate for classical music reproduction however, and often made they up for weaknesses due to inadequate cone diameter and linear displacement.

Additionally, that sharper LF amplitude roll-off leads to a more intense phase change variation about the corner frequency. 

This is where phase relationships of fundamental notes from below, and their harmonics above driver resonance can become shifted, (they are transduced out of time with each other when compared to the original energising waveform) so that when reproduced, individual instruments lose their identitifiable characteristics within the music and no longer have separatable dynamics. 

What with this loss of dynamic attack, the common driver resonance and the loss of harmonic integrity, is it any wonder that drum and bass guitar notes can sometimes become indistinguishable.

If we want dynamic reproduction we do not want the loudspeaker imposing its own tonal characteristics upon music, so, irrespective of driver size, my choice for OB driver Qes is likely to be between 0.55 and 0.35, which is quite different to the higher Q figure so often recommended by others !

Cheers ........... Graham.

[JeffB]

Thanks Graham,

That explanation and those graphs were really helpful.

[Danny Richie]

Hey Jeff,

I designed the speakers that you mentioned (LS-9's), including much of the driver design. So I'll try to answer some of your questions and give you a little perspective on how to get what you're shooting for.

So I am curious about the general question.  Can better base be obtained from multiple smaller speakers than a single bigger speaker?

Sure, and sometimes considerably.

Is the answer to this question any different for Open Baffle designs?

The principles that make the smaller driver a better choice can still apply.

I have a 15" woofer on an open baffle and from 100Hz down it is some of the best bass that I have heard.

No disrespect meant, but if what you are used to is an uncontrolled 15" woofer then there is a whole new world out there for you in low bass reproduction.

It is better than my sealed sub box with two 10s.  It is better than a ported speaker that I have with two 8s.
I believe it to be better than the Paradigm Studio 100s I demoed with four 7s.
It is better than all the Polk speakers I demoed at Fry's.

With all due respect, you really aren't comparing to anything significant.

AV123 states that a bigger woofer will have more mass, so is harder to stop, and will not as accurately handle note decay.
In the case of the 15", I can definitely imagine this affect above 100Hz, but I am not convinced it is relevant below 100Hz.

What you read was correct, but it is just the opposite as frequency increases.

The shorter wavelengths impose much less work to reproduce. Exertions are less. Stored energy is less.

As frequency decreases wavelengths get much longer and exertions needed to reproduce them get much longer as well. Stored energy and inertia are much greater. As a result the time needed to return back to a resting position is much greater.

I am also puzzled how multipe 6.5" drivers even make bass.

Actually low end extension is not a problem even with a small driver. It is all in how the driver is designed. I can give you really strong low end extension from a small driver if you don't mind really low sensitivity. I could even do it with a woofer smaller than 6", but a woofer that is smaller than 6" and that will play flat to 20Hz is going to have a peak output level of about 75 to 80db. That's not real practical. It is all about trade offs.

Now if you are using twelve 6.5" woofers to get the higher SPL needed, then you can design one to give good low end extension. Now a 30Hz Fs and 84 to 85db sensitivity is not a problem.

However, the response of most single 6.5" drivers drops off fast starting around 150Hz to 200Hz or so.

You certainly can't think of the woofers used in the LS series in terms of most 6.5" woofers. These are purpose build woofers. Even so, most 6.5" woofers designed for general applications can get you to a -3db point (ported) in the 40 to 50Hz range.

I have a hard time envisioning the responses down around 30Hz adding up.
Would such an arrangement need a contour filter to pull up the bass?

Actually they LS series is flat in output down to a really low range and cause some overloading in most rooms. There is a bass management system built into those speakers that allows the end user to reduce output in the 30 to 80Hz range to minimize the room gains in that region.

In my case, I have an 8" Visaton B200 mated with a 15" Warrior on an 18" baffle.
I am essentially lacking that hard hitting base thump of the kick drum.
I get sufficient SPL in the 100Hz to 200Hz region, but I don't think I have sufficient control in this region.

The only real "control" you are going to have is more directly associated with the compliance of the driver. This is a different issue than output level.

You can add gain as needed to maintain a flat response. Some type of transfer circuit will give you that.

Quality bass response is not in the output. Putting it into motion is not a problem. It is all about stopping the motion that will give you quality bass response. This is why the multiples of smaller drivers blow away a single large one.

With those large 15" woofers you not only have a heavy mass to sling around but no box around it to add resistance to help damp it. You are completely relying on the suspension.

For what you are doing you might want to do some reading on our new Direct Servo subs.

There is a ton of info here, including explanations from the designer and patent holder:  http://www.audiocircle.com/index.php?topic=47310.0

You can get some more info on the SW-12-16FR here: http://www.gr-research.com/index.asp?PageAction=VIEWPROD&ProdID=141

These are designed to operate in free air. They will give you the bass control you're looking for and be essentially cleaner and have better resolution and any other driver that you can find for that application.

[JeffB]

Thanks Danny, very informative.

I find this interesting

but it is just the opposite as frequency increases.

I assumed that since the driver has to move faster with increasing frequency that it would have a harder time stopping.
The Warrior 15" crossed-over at 100Hz is good, and is a little muddy at 150Hz and un-listenable at 200Hz.

I guess my biggest confusion with small drivers making base is related to Fs.
I am assuming that no matter how many drivers you use you are not going to get much base below Fs, unless maybe you use equalization.
Even then, you are going to have phase issues below Fs.

I understand the idea that one could design a special driver with a lower Fs by giving up efficiency and then getting that efficiency back by using multiple drivers.  This sounds ideal. 
I am not aware of any such drivers that are commercially available.

My thought earlier was to try four 10" Warriors, simply because it is not too expensive to try.
However, the Fs on the 10" Warrior is 65Hz vs the 15" Warrior with Fs of 36Hz.
So I am wondering if this is a reasonable thing to try. 
Or do I need a 10" driver with an Fs of 36Hz to get an equivalent low end response.

[JeffB]

What do you know a 10" woofer with an Fs=36Hz.  4Ohm though and twice the price of the Warrior.
http://www.electronixwarehouse.com/car/sgl-prod/speakers/woofers/spl/s1-series.htm