I tried it the other way round. I had the small speaker in a closed box behind, and a bigger speaker (full range) in front, each driven by a separate channel of a stereo amp. I did it to get the big speaker in a small box. It was not a true isobaric, or indeed intended as such, because the connecting chamber had to have some volume. It worked ok. The idea was that the small speaker would dictate the low frequency extension.
But if 2 (different) speakers were wired in parallel, I am not sure what would be the result.
But if 2 (different) speakers were wired in parallel, I am not sure what would be the result.
I once modeled something very similar to Midrange's idea. In my case it was to take a high-quality fullrange driver that preferred a big enclosure and make it the front of an isobaric pair in a small box, where the bass-optimised rear driver would be driven at low frequencies only. The intent was to "unload" the front driver so it could perform as though it was in a larger box. I was persuaded that a FAST style design would be a better use of the drivers.
It's easy to simulate in AkAbak, I'll take a look and see if I saved my experiments.
It's easy to simulate in AkAbak, I'll take a look and see if I saved my experiments.
Hi TJF,
there would be some criteria for estimating, whether that "asymmetric compound driver" is kind of "valid".
E.g. the maximum displaced Volume - given a certain limit for distorsion - should be about the same at both drivers IMO ...
But you have to calculate (measure) all relevant parameters of the new "Frankenstein" driver. And there are some tricky parts also.
So at least we got a name for it ...
But the question is, what do you like to achieve really ?
In the end there will arise the question "why not (find and) use a suitable (single) driver for your speaker project first of all ?"
If you just want to lower the free air resonance of the driver (not an expensive one i hope ...),
you may (at your own risc according to traditions here):
- add mass to the cone and also increase B x L to adjust Qes ... (e.g. add magnet rings the right way or change magnet)
Alternatively:
- add some mass virtually by clamping a fat capacitor in parallel to the driver terminal
- cut out some "piece of cake" shaped pieces from the suspenson (spider) to make it less stiff (but you increase Vas)
- use an active compensation filter modifying the rolloff at the low end
Of course all this needs some planning and calculation to be done if the driver you need should really be unavailable on the market ...
But no need to make a compound driver for playing such "dirty" tricks.
Compound drivers (e.g. 2x identical drivers electrically in parallel) give
- same Qt and Fr but half equivalent volume like the two constituent drivers, which allows for half size enclosure, which is nice (*)
- this is "paid" by half power efficiency as voltage sensitivity stays the same but impedance is half (=> double current and power)
Cheers
_________________
(*) Also a "push/pull" configuration may be used to make the compund driver "more symmetric" due to inward/outward movement, lowering even order distorsion this way.
All in all compound drivers can be a nice option especially in subwoofers ...
there would be some criteria for estimating, whether that "asymmetric compound driver" is kind of "valid".
E.g. the maximum displaced Volume - given a certain limit for distorsion - should be about the same at both drivers IMO ...
But you have to calculate (measure) all relevant parameters of the new "Frankenstein" driver. And there are some tricky parts also.
So at least we got a name for it ...
But the question is, what do you like to achieve really ?
In the end there will arise the question "why not (find and) use a suitable (single) driver for your speaker project first of all ?"
If you just want to lower the free air resonance of the driver (not an expensive one i hope ...),
you may (at your own risc according to traditions here):
- add mass to the cone and also increase B x L to adjust Qes ... (e.g. add magnet rings the right way or change magnet)
Alternatively:
- add some mass virtually by clamping a fat capacitor in parallel to the driver terminal
- cut out some "piece of cake" shaped pieces from the suspenson (spider) to make it less stiff (but you increase Vas)
- use an active compensation filter modifying the rolloff at the low end
Of course all this needs some planning and calculation to be done if the driver you need should really be unavailable on the market ...
But no need to make a compound driver for playing such "dirty" tricks.
Compound drivers (e.g. 2x identical drivers electrically in parallel) give
- same Qt and Fr but half equivalent volume like the two constituent drivers, which allows for half size enclosure, which is nice (*)
- this is "paid" by half power efficiency as voltage sensitivity stays the same but impedance is half (=> double current and power)
Cheers
_________________
(*) Also a "push/pull" configuration may be used to make the compund driver "more symmetric" due to inward/outward movement, lowering even order distorsion this way.
All in all compound drivers can be a nice option especially in subwoofers ...
Last edited:
There's zero benefit to be had. The best, fairytale case is that it would perform close to a single box-appropriate driver. You will not achieve improvements in distortion, power handling, low frequency extension/bandwidth, output, or any other metric without severe, crippling compromises on some other of those aspects.
Done with well matched drivers, Isobaric can be excellent. Without the matching, it's completely worthless.
Done with well matched drivers, Isobaric can be excellent. Without the matching, it's completely worthless.
There is no significant advantage to any isobaric alignment other than reduction of cabinet volume. And that really only works for large drivers with large VAS. Otherwise the reduction in volume is offset by the space required for the additional driver and the isobaric chamber. Obviously the disadvantages are numerous: need for matched drivers; loss of efficiency, added complexity and cost.
Hi John,
yes this is a very practical aspect IMO, so one won't use it in bookshelves or even normal sized stereo main speakers usually.
"Need for matched drivers" i am not so sure:
One could even (mis-) use compound in subwoofers to balance e.g. deviations in Vas (by tolerances in spider compliance) by making "averaged" compound drivers intentionally.
So compound technique could even reduce tolance (for the resulting compound drivers !), although "pairwise selection" or even changing the manufacturer are IMO favourite ways to go usually with tolerances in a specific type of driver, depending on severity of the issue ...
At LF the two constituent driver's membranes (of a compound driver) are kind of "glued together" up to a certain frequency, depending on sufficiently small volume in the coupling chamber. Thus the "matched driver" argument i am not willing to follow (as "increasing cost") here, at least not in a subwoofer application.
Whether you match drivers pairwise in a "conventional" design or select them for "balancing" also in a compound driver e.g., there is no real difference in effort IMO and of course you can do both ...
Also decrease in even order distorsion(*) can be significant with certain drivers, so the compound driver may be used in making a "higher quality driver" than it's "pre-procucts" were (!):
This is also speaking against compound always being just an increase in cost.
Of course there is always some larger manufacturing effort at the cabinet, no doubt here. But in some cases can be reduced to a "kind of gasket ring" or "inbetween plate", e.g. just holding two constituent drivers together "face to face", so even that point is a relative one due to application IMO ...
Kind Regards
___________
(*) Although often not being an audible problem. And not only in rare cases some of the most distorted subwoofers succeed in listening tests, with the "testers" or "listening experts" stating it to sound "warm" and "tight" and "dry" and "detailed" and whatnot ...
Last edited:
Hi, Thomas:
I noticed your initial query on this topic with interest, as - coincidentally - I am presently working on just such a project. That is: using two somewhat similar woofers from different manufacturers, of the same nominal size but definitely not "matched" in their parameters; in an isobaric configuration. Presently I have measured the T/S parameters of the individual drivers, and have played around with WinISD to see if isobaric would be worth considering, and I've decided it is. Pending some repair work, the next step will be to measure the parameters of the woofers mounted together as a single unit.
With consideration of the potential challenges facing this concept already posted to this discussion, I am reluctant to offer further details of my project until I have some more useful data to present. Perhaps within a week or so I can follow up with my findings, as well as an explanation of my reason for attempting this, pictures, etc. (if anyone's interested...)
Wilf
I noticed your initial query on this topic with interest, as - coincidentally - I am presently working on just such a project. That is: using two somewhat similar woofers from different manufacturers, of the same nominal size but definitely not "matched" in their parameters; in an isobaric configuration. Presently I have measured the T/S parameters of the individual drivers, and have played around with WinISD to see if isobaric would be worth considering, and I've decided it is. Pending some repair work, the next step will be to measure the parameters of the woofers mounted together as a single unit.
With consideration of the potential challenges facing this concept already posted to this discussion, I am reluctant to offer further details of my project until I have some more useful data to present. Perhaps within a week or so I can follow up with my findings, as well as an explanation of my reason for attempting this, pictures, etc. (if anyone's interested...)
Wilf
As I understand TJF's requirement, he wants to achieve a deeper overall frequency response for a given driver in a small enclosure, as if it were in a larger enclosure.
The simplest way to reduce the size of the enclosure while keeping the same response is to mount a second identical driver in "isobaric" configuration. However, their low-frequency characteristics such as Vas may still result in a larger than desirable enclosure. Replacing the "rear" or "inner" driver with one with a lower Vas may allow making the enclosure smaller than for the two high-Vas drivers.
If this is to be done without adding EQ, the parameters of the drivers must be carefully chosen so that the "assistance" provided by the rear driver increases the effective compliance of the rear driver / enclosure combination to the optimum value for the front driver. The math for this should not be significantly harder than the math required for a standard enclosure, for those with the math skills, which I don't have. I'd have to model it in AkAbak, tweak the rear driver parameters until I got a suitable result, then see if any drivers existed that would match the parameters.
The rear driver / enclosure should ideally have the same system resonant frequency as the front driver in an optimum enclosure. The temptation may arise to design the driver combination so that the "front" driver is driven below its normal resonant frequency in order to extend the bass response. As the frequency being reproduced drops, the displacement (movement) of the cone has to increase to maintain a flat frequency response. The combination of drivers should avoid driving the "front" driver beyond its linear excursion limits. The problem is acute if the front driver is a wide-range or full-range unit, which usually have small excursion limits. In this case, reverting to a standard 2-way configuration may allow achieving higher SPL.
Incidentally, it is commonly stated that the pressure between two identical drivers in an "isobaric" configuration is constant (unchanging), hence the "isobaric" term. This is incorrect. The pressure between the drivers varies by half the amount that it varies in the enclosure. Provided the volume between the drivers is much smaller than the volume of the enclosure, the effect can usually be ignored. It can cause problems if the enclosure is ported, because the effective compliance of the enclosure becomes very small at resonance, making the compliance of the inter-driver volume significant. A true" isobaric" pair of drivers would have the cones rigidly coupled together.
The simplest way to reduce the size of the enclosure while keeping the same response is to mount a second identical driver in "isobaric" configuration. However, their low-frequency characteristics such as Vas may still result in a larger than desirable enclosure. Replacing the "rear" or "inner" driver with one with a lower Vas may allow making the enclosure smaller than for the two high-Vas drivers.
If this is to be done without adding EQ, the parameters of the drivers must be carefully chosen so that the "assistance" provided by the rear driver increases the effective compliance of the rear driver / enclosure combination to the optimum value for the front driver. The math for this should not be significantly harder than the math required for a standard enclosure, for those with the math skills, which I don't have. I'd have to model it in AkAbak, tweak the rear driver parameters until I got a suitable result, then see if any drivers existed that would match the parameters.
The rear driver / enclosure should ideally have the same system resonant frequency as the front driver in an optimum enclosure. The temptation may arise to design the driver combination so that the "front" driver is driven below its normal resonant frequency in order to extend the bass response. As the frequency being reproduced drops, the displacement (movement) of the cone has to increase to maintain a flat frequency response. The combination of drivers should avoid driving the "front" driver beyond its linear excursion limits. The problem is acute if the front driver is a wide-range or full-range unit, which usually have small excursion limits. In this case, reverting to a standard 2-way configuration may allow achieving higher SPL.
Incidentally, it is commonly stated that the pressure between two identical drivers in an "isobaric" configuration is constant (unchanging), hence the "isobaric" term. This is incorrect. The pressure between the drivers varies by half the amount that it varies in the enclosure. Provided the volume between the drivers is much smaller than the volume of the enclosure, the effect can usually be ignored. It can cause problems if the enclosure is ported, because the effective compliance of the enclosure becomes very small at resonance, making the compliance of the inter-driver volume significant. A true" isobaric" pair of drivers would have the cones rigidly coupled together.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.