I think isobaric is kind of interesting, but I think Dave hits the nail on the head here.
Basically they were popular in the early 90s, when there weren't a lot of drivers that would work in small enclosures. Amplifiers began getting a LOT cheaper around 2000 or so, and that really encouraged speaker vendors to sell drivers that work in tiny boxes. Dan Wiggins from Adire Audio wrote a lot of convincing papers, arguing that xmax is basically the most important loudspeaker parameter, especially if paired with a motor that can flatten out the BL curve.
So...
1) isobaric is kinda interesting with cheap/old woofers; you can use a smaller box
2) but you can often achieve the same goal by just using a modern woofer with a small VAS
I dont believe tubelectron is planning on EQ after the cabs are built, just passive xover.
Yes @wolf_teeth : passive crossover. And I already have the consistent and ruggered Beyma 12BR70 Woofers in stock. The idea is to build a "challenger" to my 375L 3-Ways Bass-Reflex (at left), with - as possible - what I have on hand :
But it's certainly possible to build a sealed, non-isobaric enclosure with a "modern" Woofer : which one would you then recommend, Guys ?
T
With such a construction, you got to mind the SPL of the chassis. If the woofer is too loud, it makes no sense.
If the mid woofer is too loud and you use large resistors to attenuate it, the sound will suffer.
So better carefully match the drivers before you start a physical construction.
There is another point that makes it not the same to isobaric if you simply DSP a single driver: The cone of the isobaric driver only sees half the pressure difference of a single one. It is virtually stiffer.
While your amp has to deliver two times the power, it is in fact divided over two voice coils.
A DSP'd single driver's voice coil may not be able to stand the power needed to drive the cone to the desired position and the cone itself may not be strong enough to stand the forces driving it. Not unrealistic imo.
So if you add up the advantages, the lower thermal load on the magnet and voice coil and the virtually stiffer membrane may result in a better sound from the isobaric woofer construction. If we don't have another driver to choose for our speaker, I may add.
In the world of HIFI, where even the smallest difference can start religious wars, like wire sound, such a difference may be worth noting.
If the mid woofer is too loud and you use large resistors to attenuate it, the sound will suffer.
So better carefully match the drivers before you start a physical construction.
There is another point that makes it not the same to isobaric if you simply DSP a single driver: The cone of the isobaric driver only sees half the pressure difference of a single one. It is virtually stiffer.
While your amp has to deliver two times the power, it is in fact divided over two voice coils.
A DSP'd single driver's voice coil may not be able to stand the power needed to drive the cone to the desired position and the cone itself may not be strong enough to stand the forces driving it. Not unrealistic imo.
So if you add up the advantages, the lower thermal load on the magnet and voice coil and the virtually stiffer membrane may result in a better sound from the isobaric woofer construction. If we don't have another driver to choose for our speaker, I may add.
In the world of HIFI, where even the smallest difference can start religious wars, like wire sound, such a difference may be worth noting.
Yes. Here, in accordance to the manufacturer's specs we would have :
* Woofer : Beyma 12BR70 : N=93dB/W/m (for a single unit)
* Low-Mid : Beyma 8M60N : N=95dB/W/m
* High-Mid : SB Acoustic SATORI MD60N : N=94dB/W/m
* Tweeter : Beyma T2030 : N=95dB/W/m
As you can see, the sensivity between the speakers is already quite well balanced.
Moreover, the usable bandwidth for each speakers is wide enough to offer the suitable linear recovery around FCs, as recommended for a 1st order Serial/Parallel crossover, that I experimented with success on my 3-ways 375L.
So the operating conditions should be favourable at the start, for project experimentation... 😉
T
Don't forget baffle step - you're not going to get 93dB sensitivity in the bass region.
There are conflicting data, I only see 90dB for your woofer. Anyway, because of it's data it is a typical candidate for ISO. Has a small magnet and needs a very large cabinet, vented is no option at all.
Maybe just build a test cabinet of similar size and measure the result with a microphone, compared to the other driver. Even if it is called "test cabinet", it has to be air tight. I have seen to many leaky "test" builds with faulty results, so excuse me for pointig this out.
Measuring TSP is a little bit simpler and will theoretically tell you just the same. Anyway, it gives you a close idea what bass you are going to get when you crank up the test ISO build.
Maybe just build a test cabinet of similar size and measure the result with a microphone, compared to the other driver. Even if it is called "test cabinet", it has to be air tight. I have seen to many leaky "test" builds with faulty results, so excuse me for pointig this out.
Measuring TSP is a little bit simpler and will theoretically tell you just the same. Anyway, it gives you a close idea what bass you are going to get when you crank up the test ISO build.
I intended to build an isobaric car sub, since the optimum size for the woofers I had was 4.8 cubic feet, but discovered that because of the sloping top, there wasn't enough clearance for the tunnel inside. And as it turned out, it sounded/felt simply wonderful with a single driver in 2.4 cubic feet.
Really, it's a pointless configuration. It costs twice as much for drivers, requires twice the power, and the volume saving is never really half, because the second driver and tunnel occupy space. If cabinet volume is truly limited, mass-loading a driver or choosing a different driver is a better option.
Really, it's a pointless configuration. It costs twice as much for drivers, requires twice the power, and the volume saving is never really half, because the second driver and tunnel occupy space. If cabinet volume is truly limited, mass-loading a driver or choosing a different driver is a better option.
I don’t know why reducing distortion of any kind 2nd, 4th would be a bad thing. Maybe if 3rd is too high you should choose a different woofer?
Sometimes you can choose a driver with a much larger diaphragm and put two of them in a box of equivalent volume of one long throw smaller woofer. Doing this can move more air, which is the point.
I’ve built several isobaric speakers too. They have their place. They do suck up power! I don’t think they sound better, they simply are smaller. People get confused because there are two motors so they think there is more strength.
One idea is to put IB speakers in an isobaric enclosure for testing. You know, before cutting holes in the oak floor.
There is little 4th. It is known that things tend to sound best when 3rd is less than second. push-pull decreases second at the expense of increasing third flipping the relative levels.
If you have a wooferwith symetrica in-out (ie a goodwoofer) then nothing changes.
dave
Doesn’t the 3rd order distortion stay at the same level? It’s the 2nd that gets reduced. I admit though that after the 2nd is reduced it could be a ‘wow’ experience as the 3rd appears more obvious. I guess that could be bad.
‘Mr Bond, this is Tweady Midelle Goodwoofer.’ When I read “goodwoofer” it cracked me up. 😂
Finally, the Isobaric chamber will look like this, with 2x12" Beyma 12BR70 :
The matching expected performances should be as below
The external dimensions of the enclosure are still HxWxD = 900x400x360mm, 4-ways isobaric, with a 8" Low-Midrange loaded by a 9.2L sealed enclosure.
Wait and See...
T
The matching expected performances should be as below
The external dimensions of the enclosure are still HxWxD = 900x400x360mm, 4-ways isobaric, with a 8" Low-Midrange loaded by a 9.2L sealed enclosure.
Wait and See...
T
This looks like a good concept. Make it 4-way active with 5 amps and you will have a hell of a box.
Passive will become much too expensive and complicated, with a never satisfying result.
PS if you want an idea for amps, 3x TPA3255 for low and 2xLM3886 would be my first choice. You find such D with A/B amp combination in some high class Studio monitors. Beside from the better end result, calculate cost for active and passive and you will be surprised that better will be less expensive. Do not forget the parts that will not work, but which you may have bought trying out combinations.
Even size will be about the same, you will need a lot of space for all these passive parts.
You will not need a single resistor to match levels in the active version. That alone will make it sound superior.
If you are hard core passive and have a fantastic amp already, only do the 3 woofers active.
Passive will become much too expensive and complicated, with a never satisfying result.
PS if you want an idea for amps, 3x TPA3255 for low and 2xLM3886 would be my first choice. You find such D with A/B amp combination in some high class Studio monitors. Beside from the better end result, calculate cost for active and passive and you will be surprised that better will be less expensive. Do not forget the parts that will not work, but which you may have bought trying out combinations.
Even size will be about the same, you will need a lot of space for all these passive parts.
You will not need a single resistor to match levels in the active version. That alone will make it sound superior.
If you are hard core passive and have a fantastic amp already, only do the 3 woofers active.
Last edited:
Thanks @Turbowatch2 !
I do have some active crossovers :
And enough amps to drive each speaker - DIY or not, Tube or not...
And I also have a bunch of R, L and C accumulated by donations or with the years and experiments passing by...
But as you said it, I am rather a "hard core passive" crossover nerd !
But yes, the question of active crossover arose when I worked on my 375L enclosures :
Before going that "active" route, I first experimented many classical passive crossovers variations, and I finally came up with this very simple crossover solution below - that proved to be the best souding by far - of all my tests :
So I will transpose that principle in a 4-ways configuration, expecting to check that it will operate as successfully as the 3-ways version :
T
I do have some active crossovers :
- one 2/3/4 ways 24dB/Oct. BEHRINGER CX3400 SuperX Pro
- one 3 ways 12dB/Oct. IMG 321 SW
And enough amps to drive each speaker - DIY or not, Tube or not...
And I also have a bunch of R, L and C accumulated by donations or with the years and experiments passing by...
But as you said it, I am rather a "hard core passive" crossover nerd !
But yes, the question of active crossover arose when I worked on my 375L enclosures :
Before going that "active" route, I first experimented many classical passive crossovers variations, and I finally came up with this very simple crossover solution below - that proved to be the best souding by far - of all my tests :
So I will transpose that principle in a 4-ways configuration, expecting to check that it will operate as successfully as the 3-ways version :
T
I have not the slightest ambition to dump my a## into series crossover construction. I'm good in doing conventional 12 and 18dB stuff.
I get two channels of a very good amps for about 40$. Very good means better than 95% of amps that existed in 1990 and even later.
If I buy from Mr. Mundorf, who has startet the 2000% price rise in passive x-over parts, I get a medium size capacitor and a small coil for that money. What made speaker DIYS so expensive that it has come out of fashion for young people that practiced it in my youth. Today old men build speakers. He robbed them of the option to learn something practical that is revarding and educative at the same time. It is something personal that not anyone can understand.
Going active is some kind of self defense. At least I don't know the Chinese that build my amp components.
If you are using conventional active crossover, I understand you, but I use DSP's. Which gives me an endless number of virtual crossover parts at a mouse click.
Once you have used that and compared the result to passive, there is no way back.
If you could still do a high quality 2-way crossover for 20% of the chassis price, things would be different.
I get two channels of a very good amps for about 40$. Very good means better than 95% of amps that existed in 1990 and even later.
If I buy from Mr. Mundorf, who has startet the 2000% price rise in passive x-over parts, I get a medium size capacitor and a small coil for that money. What made speaker DIYS so expensive that it has come out of fashion for young people that practiced it in my youth. Today old men build speakers. He robbed them of the option to learn something practical that is revarding and educative at the same time. It is something personal that not anyone can understand.
Going active is some kind of self defense. At least I don't know the Chinese that build my amp components.
If you are using conventional active crossover, I understand you, but I use DSP's. Which gives me an endless number of virtual crossover parts at a mouse click.
Once you have used that and compared the result to passive, there is no way back.
If you could still do a high quality 2-way crossover for 20% of the chassis price, things would be different.
I'd make the isobaric section as a 0.5 woofer. See what a big coil (10 mH) can do. The Beyma midrange can cope with bass without distortion, it's a pro...
But maybe since the woofers have zero gain respect to one and impedance is 8 Ω/2 ...
I say that because you insist with that odd crossover, but a practical and 'easy' parallel crossover is better ( to begin,)
But maybe since the woofers have zero gain respect to one and impedance is 8 Ω/2 ...
I say that because you insist with that odd crossover, but a practical and 'easy' parallel crossover is better ( to begin,)
A matching series crossover needs less parts and because of that should sound better than a conventionel one.
I hardly understand them and never tried to build one.
Anyway, if such a crossover is better because of less passive components in the signal path, the best crossover is one without any passive parts. Which leads you directly to active and a DSP. Since there isn't any music that has not been converted to bits, my problem with digital in the crossover is nill.
I hardly understand them and never tried to build one.
Anyway, if such a crossover is better because of less passive components in the signal path, the best crossover is one without any passive parts. Which leads you directly to active and a DSP. Since there isn't any music that has not been converted to bits, my problem with digital in the crossover is nill.
This is an unusual, odd-design passive crossover, yes.
I first experimented the filtering on my 375L with conventional designs in 12dB/Oct. and locally 18 and 24dB/Oct.
These crossovers indeed looked like nuclear plants... 😆 But, OK : if it is "for the good cause", why not ?
Then I thought why not test this ridiculously simplistic serial-parallel configuration ? (inspired from the works of Mr. Bakegaard, from B&O).
The results were clearly at the advantage ot the 612SP : "lively against lifeless", if I can say so.
Amazing audible difference : if I did not tested it, I would not have believed it...
So I called together some audiophile friends for a blind listening session, to give their advice between the two crossover solutions.
Same tests and same results : the superiority of the "odd simplistic" 612SP proved unanimously the best.
My Wife and Daughter - not Audiophile in any way, but both musicians - told me that both were balanced, but the superiority of the 612SP laid in the "blend" and "synchro" of the 3 loudspeakers vs. the conventional 12dB/Oct. Their words can be interpreted as homogeneity of the timbre and precision of the transients, IMO.
In fact, this crossover seems to be a 1st order one, but with slopes starting at 6dB/Oct. and then tending to reach 12dB/Oct. , all speakers connected in polarity.
Below, a 3-Ways Xover / 1st Order (6dB/Oct.) / 250Hz - 4kHz : at LEFT, classical configuration ; at RIGHT serial-parrallel 612SP configuration :
For more details, see this website (sorry, in French) :
http://jimbee.over-blog.com/page-5371974.html
From there, why not test this 612SP crossover architecture in a 4-ways version, before testing other more conventional or more sophisticated solutions in comparison ?
T
I first experimented the filtering on my 375L with conventional designs in 12dB/Oct. and locally 18 and 24dB/Oct.
These crossovers indeed looked like nuclear plants... 😆 But, OK : if it is "for the good cause", why not ?
Then I thought why not test this ridiculously simplistic serial-parallel configuration ? (inspired from the works of Mr. Bakegaard, from B&O).
- One loudspeaker with the conventional crossover (12dB/Oct.), and the other with the 612SP crossover.
- A/B mono tests, and Stereo tests.
- L and R then R and L. (I mean : shift of the crossovers between the two enclosures).
The results were clearly at the advantage ot the 612SP : "lively against lifeless", if I can say so.
Amazing audible difference : if I did not tested it, I would not have believed it...
So I called together some audiophile friends for a blind listening session, to give their advice between the two crossover solutions.
Same tests and same results : the superiority of the "odd simplistic" 612SP proved unanimously the best.
My Wife and Daughter - not Audiophile in any way, but both musicians - told me that both were balanced, but the superiority of the 612SP laid in the "blend" and "synchro" of the 3 loudspeakers vs. the conventional 12dB/Oct. Their words can be interpreted as homogeneity of the timbre and precision of the transients, IMO.
In fact, this crossover seems to be a 1st order one, but with slopes starting at 6dB/Oct. and then tending to reach 12dB/Oct. , all speakers connected in polarity.
Below, a 3-Ways Xover / 1st Order (6dB/Oct.) / 250Hz - 4kHz : at LEFT, classical configuration ; at RIGHT serial-parrallel 612SP configuration :
For more details, see this website (sorry, in French) :
http://jimbee.over-blog.com/page-5371974.html
From there, why not test this 612SP crossover architecture in a 4-ways version, before testing other more conventional or more sophisticated solutions in comparison ?
T
This filter with changing slopes seems interesting but did it have really advantage against ordinary filters if counting drivers impedance curves with resonances and drivers current non-linearities?
It is interesting to simulate this filter with real driver data on Vituixcad.
It is interesting to simulate this filter with real driver data on Vituixcad.