The idea of using multiple small fullranges instead of a single midwoofer has been increasingly haunting me recently. What problems and potential advantages do you see for a practical design case of four 3” drivers tightly colocated ? Some of them come in square shape with a side length of 80mm. So colocating them next to each other would give a virtual square midwoofer of 160mm x 160mm size, so from the outer dimensions roughly equivalent to a single 6” driver. I would think of connecting them in paralel in pairs and amplify separately. If my assumptions are right that would mean a 12dB efficiency uplift compared to a single driver FR. Then using some low-pass FR DSP shaping I could preserve the bass efficiency and then above 200Hz adjust it to the target upper range efficiency. I can see quite a few advantages of such design but have never seen anything similar. Maybe there are good reasons also in terms in acoustic quality why it is not practised ?
And a related question - what would be a good driver for this application. In my research I have initially selected the attached Peerless but when studied its parameters I got doubts. I do not have too much experience in TS parameters and different driver design philosophies so I would greatly appreciate some enlightement here. What are the key differences between high Qts drivers like this Peerless and less stretched ones like e.g. the attached TangBand that is also in the game ? What is the intended application of both ? It surprises me that TangBand has a few times smaller linear excursion than Peerless - is it again some consequence of different compromises, design philosophy and/or intended application ? It seems Peerless might not be intended for a closed box that I am considering here – the Q enclosure is then forced to be high and the internal volume large ?
Pawel
And a related question - what would be a good driver for this application. In my research I have initially selected the attached Peerless but when studied its parameters I got doubts. I do not have too much experience in TS parameters and different driver design philosophies so I would greatly appreciate some enlightement here. What are the key differences between high Qts drivers like this Peerless and less stretched ones like e.g. the attached TangBand that is also in the game ? What is the intended application of both ? It surprises me that TangBand has a few times smaller linear excursion than Peerless - is it again some consequence of different compromises, design philosophy and/or intended application ? It seems Peerless might not be intended for a closed box that I am considering here – the Q enclosure is then forced to be high and the internal volume large ?
Pawel
Currently a work in progress, in big and small versions. One problem would be beaming, but I think I can get around that. 4 ridged light cones are going to be better than one big one, the cone brake up of a small cone but the surface of a larger on, and possibly more efficiency.
I would appreciate some insight from speaker designers who could provide some hints how to understand and translate the T-S parameters into practice.
I am considering a few alternatives for a multiple 3" fullrange design and do not feel comfortable with quite striking differences in their parameters and enclosure design implications. Let’s take two drivers that I have been comparing recently and their parameters:
Peerless 830986 Tangband W3-1053SC
Resonance Frequency fs 105 Hz 100
Total Q factor Qts 0.76 0,52
Equivalent volume Vas 0.94 ltrs 1,77
Sensitivity (1W/1m) 83.2dB 87dB
Maximum Linear Excursion 2.1mm 0,75 mm
Calculated closed box volume Q = 0.8 8,7 lts (Peerless) , 1,29 lts (TangBand) !
The key parameters for me are:
Peerless has significantly higher max linear excursion but that actually does not translate into SPL in bass but maybe it has advantages for the expected level of harmonic distortions or just to the contrary ?
And can anything about estimated distortions level be inferred from the TS parameters or it can only be measured ?
I am considering a few alternatives for a multiple 3" fullrange design and do not feel comfortable with quite striking differences in their parameters and enclosure design implications. Let’s take two drivers that I have been comparing recently and their parameters:
Peerless 830986 Tangband W3-1053SC
Resonance Frequency fs 105 Hz 100
Total Q factor Qts 0.76 0,52
Equivalent volume Vas 0.94 ltrs 1,77
Sensitivity (1W/1m) 83.2dB 87dB
Maximum Linear Excursion 2.1mm 0,75 mm
Calculated closed box volume Q = 0.8 8,7 lts (Peerless) , 1,29 lts (TangBand) !
The key parameters for me are:
- SPL at low frequencies, I chose 80 Hz as a reference frequency
- Level of harmonic distortions
- Size of enclosure and ability to work well in a closed box
Peerless has significantly higher max linear excursion but that actually does not translate into SPL in bass but maybe it has advantages for the expected level of harmonic distortions or just to the contrary ?
And can anything about estimated distortions level be inferred from the TS parameters or it can only be measured ?
Fs of both drivers are very close, but TB has almost double Vas. That means moving mass of TB is almost half of the Peerless. Also, much smaller Xmax of the TB indicates better utilisation of available magnet flux in the gap.
Together, that gives lower Qts and higher efficiency for TB.
When you put the driver in a closed box, Fs and Qts will rise. Since Peerless has Qts=0,76 and your goal is Q in the box =0,8 that leaves very small margin for Q to rise. Hence you end up with such big Vb compared to Vas for the Peerless.
Take bigger Q for the box, say 1 - 1,2 and you will get smaller Vb.
Looks like Peerless aim is to achieve good Xmax, at expense of efficiency.
At my opinion TB, with lower Qts, is better choice for ported box.
Higher Xmax should give lower distortion.
As far as I know, they have to be measured.
Using multiples of small (4") fullrange drivers instead of one larger midwoofer is bad idea. It's not about cone surface area, multiples of small fullrange drivers can have larger surface area, that is completely beyond the point. Single 4" fullrange driver have significant distortion below 200Hz, terrible below 100Hz, and below, distortion is shooting up. This is fact. Almost all small drivers behave this way because of the design limitation in excursion.
See Timothy Feleppa's pages if you do not believe me.
http://feleppa.com.au/speakermeasmid.html
Putting multiples of the same small fullrange drivers together will not remove this distortion, it will still be there. Signal is divided between many drivers, granted, they receive less signal, but the distortion is still there.
One mid woofer (6-10") designed well to cover 30Hz to say 1000Hz will have significantly less distortion. Period.
All that nonsense about small drivers having fast bass is nothing but nonsense. Simple as that. Bass is not fast or slow, its either well executed or not.
Now paralleling multiple small driver as midranges, each used in its best fr range will bring benefits, if done well. Like lower distortion.
See Timothy Feleppa's pages if you do not believe me.
http://feleppa.com.au/speakermeasmid.html
Putting multiples of the same small fullrange drivers together will not remove this distortion, it will still be there. Signal is divided between many drivers, granted, they receive less signal, but the distortion is still there.
One mid woofer (6-10") designed well to cover 30Hz to say 1000Hz will have significantly less distortion. Period.
All that nonsense about small drivers having fast bass is nothing but nonsense. Simple as that. Bass is not fast or slow, its either well executed or not.
Now paralleling multiple small driver as midranges, each used in its best fr range will bring benefits, if done well. Like lower distortion.
Last edited:
Any opinions are greatly appreciated 🙂 And yes, that's about the only real negative characteristic about this design that I am also a bit worried about.
But I did study the Timothy Feleppa's pages (very, very helpful BTW) and do not have the same conclusions as you. Yes, for the <200Hz region the small drivers shoot up wth distortion. But in my understanding of physics that should be, contrary to what you said, dependent on the number of drivers. They get less power - they distort less.
For the >200Hz region good 3" can be better than most big bros (but much less of them at Timothy Feleppa's pages).
As for the fast bass, hm, yes, I would not say there is actually "fast bass". But there is more/less dynamic and finely texturized bass that relates to nothing in the low frequencies. The lows are just the bass energy/support (that should not be missing clearly), its colour and overall sound impression is created but what is going on much higher up.
Transients for bass are one of the most important and valuable overall in my opinion.
In the trade-off low bass distortion vs transients fidelity I am both hands for the latter. But we will see, I should know in a few weeks and then I will let know here. Unless the results will not support my case 🙂
But I did study the Timothy Feleppa's pages (very, very helpful BTW) and do not have the same conclusions as you. Yes, for the <200Hz region the small drivers shoot up wth distortion. But in my understanding of physics that should be, contrary to what you said, dependent on the number of drivers. They get less power - they distort less.
For the >200Hz region good 3" can be better than most big bros (but much less of them at Timothy Feleppa's pages).
As for the fast bass, hm, yes, I would not say there is actually "fast bass". But there is more/less dynamic and finely texturized bass that relates to nothing in the low frequencies. The lows are just the bass energy/support (that should not be missing clearly), its colour and overall sound impression is created but what is going on much higher up.
Transients for bass are one of the most important and valuable overall in my opinion.
In the trade-off low bass distortion vs transients fidelity I am both hands for the latter. But we will see, I should know in a few weeks and then I will let know here. Unless the results will not support my case 🙂
It depends but in general Adason is correct. Even thought there will be lower excursion per driver for the multiple smaller driver setup it will likely produce more distortion for a given SPL than a single larger driver of similar total Sd. A larger driver will often have more Xmax capability, and/or for the same Xmax may have less compliance variation with cone position, so less distortion. At low frequencies, a larger driver moving less is almost always better, though it may require a larger enclosure.
Also, at higher frequency the off axis response is always worse when you use multiple small drivers compared to a single larger one of comparable size. You can model this with e.g. TheEdge.
I was looking for such bits of info and there is another thread of me focused directly on that. The conclusion there has been so far that it is just the spacing between the drivers in their overall physical placement that counts.
I would like to avoid just mechanical relying on results of some tools. Do you know what causes this difference visible in TheEdge ? There will be slight differences in the shape of the wavefront because of the different cone shapes that create the waves, in particular in the case of multiple small ones there will be more of some space irregularities in the overall shape that creates the sound wave but why it would have impact just on the off-axis reponse ? I can easily imagine some FR irregularities in this case but why would they apply only off-axis ?
This idea is clearly non-mainstream but has potentially some strong advantages. BTW, if higher distortions (that's a fact) are such a deal-breaker here I am curious if strong proponents of their overruling importance have in practice heard the isolated effect of increased harmonic distortions at < 200 Hz ? That might be a difference in our approach, my aim is not minimizing these distortions my aim is achieve better sound fidelity 🙂 To have low distortions would be nice too, but this is not my primary goal here.
BTW, I have not said I am not taking into account putting an 8" in the bottom - in fact that is one of the options that is included in this contest 🙂 Maybe I did not use the right word in the title, the focus is not on that if it such multiple fullranges can play well in low bass, it is more a question how such a concept compares to a 6" woofer when applied in the >= 200Hz region. Bcs I have not seen such a design at all, neither accompanied by a big cone.
So it is more in the direction of Charlie's comments - are there some other disadvantages when not applied in low bass of such a design that it is not popular ? Or just more complicated, potentially more costly ?
BTW, I have not said I am not taking into account putting an 8" in the bottom - in fact that is one of the options that is included in this contest 🙂 Maybe I did not use the right word in the title, the focus is not on that if it such multiple fullranges can play well in low bass, it is more a question how such a concept compares to a 6" woofer when applied in the >= 200Hz region. Bcs I have not seen such a design at all, neither accompanied by a big cone.
So it is more in the direction of Charlie's comments - are there some other disadvantages when not applied in low bass of such a design that it is not popular ? Or just more complicated, potentially more costly ?
Last edited:
A loudspeaker is the worst part of the signal chain in terms of distortion. IMO an important goal for designing a good loudspeaker is to aim for as low distortion as possible as a way to compensate for this aspect of transducer behavior. If you claim to be after "better sound fidelity" then you should probably be thinking about this issue. You mention full range drivers... not to lump them all into one pot, but did you happen to see this review :
https://www.audiosciencereview.com/...research-lgk-2-0-speaker-review-a-joke.34783/
This is a fullrange driver system that uses a single 3" driver. Why did the reviewer call this speaker a "joke"? Well, check out the measured distortion:
It's not just that the distortion is "a little higher", it gets really, really, REALLY BAD! 86dB is not all that loud. The "rising" distortion should basically rule out using this driver below about 500Hz. In general, a fullrange driver is a lot of compromises that in the end don't really do much of anything well compared to multiway systems, and only offers coincident reproduction (meh). A 4" fullranger will likely have similar problems, and only the frequency below which the distortion is rising will be a bit lower (e.g. 200Hz).
And yes you can "hear" the effects of distortion below 200Hz. Up to a couple of percent it is not too bad, but once you pass about 2-3% the sound will be muddy, smeared, with less details, etc. and the more distortion present the worse this will become. Remember, HD always occurs at higher frequencies than the fundamental. The ear is more sensitive to these higher frequencies components compared to the fundamental when the fundamental is below 200Hz because of the relationship between acoustic power (SPL) and perceived loudness:
https://en.wikipedia.org/wiki/Loudness
This fact is pointed out in the left image above by the reviewer.
I tend to agree that distortion audibility has limits. In any case will the smaller drivers be pushed that hard? Some calculations to do for each case..
There comes a point where lobing is significant for a single driver, this point will be similar for the combined array.. except that breakup will be less of an issue for the multiple drivers.
I agree that breakup will be less of an issue for multiple small drivers, because for the small driver breakup will happen higher in frequency and will probably be better controlled (it may even be mostly absent).
But lower in frequency, let's say at "middle" frequencies to avoid confusion with "low" frequencies, the array of smaller drivers essentially must be worse in terms of directivity. This is because, for any given area, the smallest space that it can be fit into is described by a circle, e.g. a single driver of that area (meaning Sd). So by definition, the multiple smaller drivers must take up a wider space on the baffle. This is because the frame, being larger than the cone, makes it impossible to place the radiating cone area next to the others - there will be a gap even if the frames are touching. This is not doing to double the effective diameter of the radiating area, but it might increase it by 20% to 30%, and this is enough to make the directivity worse (e.g. the pattern will be narrower because the sources are more spread out).
I think it's up to the designer to decide if these things are important to him/her. In any case it is good to know about the issue(s) involved. It's just a hobby after all.
Just last night on my walk i was imagining a Hardwood class wall-hugging ML-TL with 3 or 4 MA CHN-50. A 3” driver. And i have an on-going project for a standmount 3 driver trapezoidal Mar_Ken with 3 x Alpair 6.2p.
So i do not consider this idea bad, as with any loudspeaker it will have compromises.
Wired in series with a big shunt cap to roll off the top of all but one of the drivers.
Personally i would pay no attention to what Charlie and Adason suggest. There are 3 and 4” FRs that produce perfectly acceptable LF, they won’t go real low, but given that even one 4” in the right box often evokes the comment “that bas sis coming from a 4”?
dave
So i do not consider this idea bad, as with any loudspeaker it will have compromises.
Wired in series with a big shunt cap to roll off the top of all but one of the drivers.
Personally i would pay no attention to what Charlie and Adason suggest. There are 3 and 4” FRs that produce perfectly acceptable LF, they won’t go real low, but given that even one 4” in the right box often evokes the comment “that bas sis coming from a 4”?
dave
I'm not seeing the point of making a connection to the area of the radiating surface rather than the radiating area of the array. It goes without saying that if you spread it out over a wider area on the baffle you will change directivity... but this should not be the default position for merely comparing the two styles.
Last edited:
Using multiple drivers in a generalized concept.
Has the usual benefits,
Increase Sensitivity, Increase thermal power, lower distortion.
as with anything taking trade offs/ limitations
into consideration for a successful end design.
Far as the directivity / dispersion issues.
It has appeared more often in live sound.
And I have seen many names to describe it.
Such as V Baffle, Splayed baffle, Crossfire , 90 degree
be nice to know a correct term if any.
I have seen various splay of 90 to 60 degrees
or basically 45 to 30 degree baffles
Anyways it has been used to provide impressive horizontal dispersion
for 4x mounted vertical and horizontal drivers.
Example below of Yorkville midrange/high frequency array.
Has the usual benefits,
Increase Sensitivity, Increase thermal power, lower distortion.
as with anything taking trade offs/ limitations
into consideration for a successful end design.
Far as the directivity / dispersion issues.
It has appeared more often in live sound.
And I have seen many names to describe it.
Such as V Baffle, Splayed baffle, Crossfire , 90 degree
be nice to know a correct term if any.
I have seen various splay of 90 to 60 degrees
or basically 45 to 30 degree baffles
Anyways it has been used to provide impressive horizontal dispersion
for 4x mounted vertical and horizontal drivers.
Example below of Yorkville midrange/high frequency array.
Last edited:
The initial post begins with the sentence:
That specific sentence formed the basis for my comments and why I made the comparison.
Ok. So..
Means Sd, not source size, and..
..means dispersion, not an off-axis inconsistency...
Gents, we all agree that there are benefits when multiple drivers are used. Not just lower distortion.
But this is assumed when these drivers are used in their normal optimal range.
Please use proper tool for the job.
Btw, i got 200 small 4" fullrange drivers aura from pe blowout sale and I have extensively used them. So i am not making stuff up.
But this is assumed when these drivers are used in their normal optimal range.
Please use proper tool for the job.
Btw, i got 200 small 4" fullrange drivers aura from pe blowout sale and I have extensively used them. So i am not making stuff up.