Isobaric driver orientation: which is best?

I am designing an isobaric (isobarik?) vented sub and would really like some input on the driver configuration. The typical gasket to gasket 'clam shell' driver mounting would certainly be easier to build but wouldn't orienting them front to back and in-phase yield better sound? It just seems that having the sound radiate from the front of the cone directly into the room would be more effective.

Also, when designing the enclosure, if the drivers are mounted in-phase will I have to make any allowances for the small front chamber?

Thanks
peter
 
Clamshell style is certainly easier to build and also much less likely to have problems. They're also usually smaller, because you only need a single enclosure. Polar alignments frequenly have standing wave problems, while clamshell alignments are no different from normal enclosures. However, clamshell designs are unlistenable for a number of reasons outside of the low frequencies. The design still works well for a subwoofer. Just not for anything else.

What drivers do you want to use for this? Efficiency is always a problem with isobaric designs, and not all drivers are suited to the job.
 
clamshell is best ..... the others sound worse and will kill your drivers quicker.

in the clamshell config, you will get nothing but bass from the speaker all the rest gets cancelled.

also, in the clamshell config, you dont get a higher spl because there is 2 drivers, infact you get less. The output will be -6dB when compared to a single driver so you will need more power.

dont use a ported box ...... it's just not worth the trouble for an isobarik speaker.

[Edited by AudioFreak on 11-20-2001 at 10:45 PM]
 
WARNING!

How many watts will you be driving your Isobaric-coupled woofers with? You should either change the design to a sealed box or use fewer watts. My Sledgehammer subwoofer used a ported box with clamshell woofers, back before I knew better. It sounds okay, but I found that it's easy to overdrive the woofers to the point where the voice coil former strikes the backplate, possibly damaging my woofers. Also, please make sure your box is the right size. My Sledgehammer subwoofer (soon to be replaced by the new American Thunder subwoofer) has a box that is 1.5 times the size it should be, and I think I may be sacrificing output and flat frequency response. This is my other reason for building my new subwoofer.

Either sealed box or fewer watts.
 
Saying that an isobaric speaker is 6dB down compared to a single driver can be misleading. Acoustically an isobaric speaker is 3dB down on a single driver, but that’s only half the story, the output will change from that depending on the connections and amp impedance.

Let’s say the amp has low output impedance and can deliver twice the power into a 4ohm load than it can into an 8ohm load, a perfect voltage amplifier, (usually not far from the truth). Then the output will be about the same if the drivers are wired in parallel, (the amps output will have doubled to make up for the 3dB loss). If the drivers are wired in series, (16ohm, not common) the amps output will have halved and the output dropped by 6dB. That is if the volume is not altered.

Isobarics are not the first choice for efficiency! They are small, generally high quality, low frequency speakers.

Have I missed something? I’m sure you’ll let me know:)

Regards WALKER
 
AudioFreak said:
clamshell is best ..... the others sound worse and will kill your drivers quicker.

i thought the only diff. was that in clamshell, any movement irregularities in a particular model's suspension are canceled since one is going "backwards" while the other is going "forwards" (and vice versa)- and of course, a smaller/tighter air piston in clams provides for less chance of uncoupling (iso = equal, baric = pressure).

AudioFreak said:
in the clamshell config, you will get nothing but bass from the speaker all the rest gets cancelled.

ummm, i'm not sure about this. one would still need a x-over. i think you're thinking of a clamshell isobaric in a 4th order bandpass like this one:

[IMGDEAD]http://www.hakalugi.net/misc/bpiso.GIF[/IMGDEAD]

AudioFreak said:
also, in the clamshell config, you dont get a higher spl because there is 2 drivers, infact you get less. The output will be -6dB when compared to a single driver so you will need more power.

but if you're using 2 drivers and wire them (or their VC's) in parallel, one can often get more power out of you amp (if it's stable) so the -6db (i thought it was -3db?) would be a worst case.--- [edit] yes, just saw walkers post, it should be 'the same' volume as one driver if your amp can double in ouput with the drivers wired in parallel. so if you amp can hang there is NO loss in output.

AudioFreak said:
dont use a ported box ...... it's just not worth the trouble for an isobarik speaker.

i'd build one if the individual driver(s) would be better suited for a ported enclosure. just think of it as a ported box that is 1/2 the size that needs twice the power.



[Edited by hakalugi on 12-04-2001 at 02:54 PM]
 
Re: WARNING!

BAM said:
How many watts will you be driving your Isobaric-coupled woofers with? You should either change the design to a sealed box or use fewer watts. My Sledgehammer subwoofer used a ported box with clamshell woofers, back before I knew better. It sounds okay, but I found that it's easy to overdrive the woofers to the point where the voice coil former strikes the backplate, possibly damaging my woofers. Also, please make sure your box is the right size. My Sledgehammer subwoofer (soon to be replaced by the new American Thunder subwoofer) has a box that is 1.5 times the size it should be, and I think I may be sacrificing output and flat frequency response. This is my other reason for building my new subwoofer.

Either sealed box or fewer watts.

i think you're mixing things here.

the damage/bottoming out are not caused by the isobaric configuration. (going isobaric only alters the Vas, Z and Pmax of the sub system).

if your box is too big, it's too big. i'm not sure if it was too big to begin with, but if it was a stock box and you added a second (identical) woofer in clamshell... the box is now definitely too big ;)

anyways, an isobaric design should be HARDER to bottom out than a traditional ported box because:

1) i takes twice the power
2) if it's clamshell, the workings of the spider and surround at it's limits are supported by the nature of the suspension in the 'other' driver at the opposite extreme. (ie: going 'non linear' or out of the gap, a driver might be 'better' going out than 'back'- but if they're face to face, you've always got one driver going 'out')

it the VC was smacking, sounds like you needed a high-pass subsonic filter or need to recalc. your port tuning frequency- you're prob. unloading the sub.
 
Dave said:
Won't mounting the drivers face to face (clamshell) provide some degree of distortion cancelation?

if the distortion is caused by suspension induced anomolies at the extremes of it's travel, then yes- which is why i mention the clam being better than the 'in-phase' tunnel type- there, there's no face-2-face suspension canceling.

but understand- when face-2-face, they're wired out of phase, so that act as 1 piston together- so nothing 'acoustically' is cancelled.
 
quote:
--------------------------------------------------------------------------------
Originally posted by AudioFreak
clamshell is best ..... the others sound worse and will kill your drivers quicker.
--------------------------------------------------------------------------------

i thought the only diff. was that in clamshell, any movement irregularities in a particular model's suspension are canceled since one is going "backwards" while the other is going "forwards" (and vice versa)- and of course, a smaller/tighter air piston in clams provides for less chance of uncoupling (iso = equal, baric = pressure).

that isn't the only difference .... Read http://www.jlaudio.com/tutorials/isobarik/

quote:
--------------------------------------------------------------------------------
Originally posted by AudioFreak
in the clamshell config, you will get nothing but bass from the speaker all the rest gets cancelled.
--------------------------------------------------------------------------------

ummm, i'm not sure about this. one would still need a x-over. i think you're thinking of a clamshell isobaric in a 4th order bandpass like this one:

No, I'm not thinking of a bandpass enclosure I was infact thinking about any clamshell isobarik speaker. Yeah, you'll still need an x-over so the sound is cleaner but frequencies higher than bass cannot be reproduced by a speaker system that does not have a driver openly facing outward. The clamshell config has a second driver in the way of the driver which is facing outward and hence cant produce anything above bass.


quote:
--------------------------------------------------------------------------------
Originally posted by AudioFreak
also, in the clamshell config, you dont get a higher spl because there is 2 drivers, infact you get less. The output will be -6dB when compared to a single driver so you will need more power.
--------------------------------------------------------------------------------

but if you're using 2 drivers and wire them (or their VC's) in parallel, one can often get more power out of you amp (if it's stable) so the -6db (i thought it was -3db?) would be a worst case.--- [edit] yes, just saw walkers post, it should be 'the same' volume as one driver if your amp can double in ouput with the drivers wired in parallel. so if you amp can hang there is NO loss in output.

Yep, you got it... I stated above that i got it wrong :)

quote:
--------------------------------------------------------------------------------
Originally posted by AudioFreak
dont use a ported box ...... it's just not worth the trouble for an isobarik speaker.
--------------------------------------------------------------------------------

i'd build one if the individual driver(s) would be better suited for a ported enclosure. just think of it as a ported box that is 1/2 the size that needs twice the power.

It aint that easy :) you'll probably have to do a few simulations to ensure that you get the best combination of SPL for your Xmax and anyway, isobaric does better in a sealed enclosure.
 
Some comments about this topic, according to my experience

Sorry, I don't know how to insert pictures in the message, so I've attached a small Excel file with some draft!

Driver Isobaric configuration

- configuring the cones in opposite physical shape (i.e. clamshell) you should have a 'mechanical' reduction of even order distortion, assuming you use identical speakers. Honestly, I never made comparative tests, so I can't say if this theory brings effective 'listening' results!

- clamshell configuration provides the bigger distance between magnets (they could interfer if too close)

- Clamshell is not indicated if you want to deliver freq. above 100-150hz max. (frame and magnet interferences with radiation)

- A rule of thumb in isobarik configuration is to keep the 'shared' air volume as small as possible. In clamshell this air volume is the 'double cone' generated by diaphgram of the 2 drivers.
So if you choose config. like B or C (see file) try to eliminate dead volumes in the isobarik box.

- If you plan high power driving for long time (i.e. PA or disco use) the clamshell configuration allows a better cooling of the coils: both magnet faces a big air volume. In other configurations (i.e. Linn) one of the magnets is in the small isobarik chamber, while the other is in abigger volume. This means their temp will be different.
(VC temperature changes the speaker parameters, i.e voice coil resistance and so Q factors etc. In the worst case your system will work with 2 drivers that are not identical anymore)


Excursion plot behaviour:

Sealed box: increasing with decrease of freq.

Vented box: similar to sealed, but with a minimum at the system resonance frequency

Bandpass 4th order (or coupled cavity) : like vented box, minimum at RS of the reflex box

Bandpass 6th order (i.e Bose's acoustimass) : 2 minimum at RS of the 2 reflex tuning

The real values for excursion must be calculated for the actual system, they depend on several parameters (box volume, speaker, quantity of filler etc.). There are softwares who do nicely the job.
 

Attachments

  • esc.txt
    24.5 KB · Views: 85