LOL!! OK Charlie - give me about an hour - I will need to have some food, the sub is still off though - so that makes life a little easier - will report back in about an hour or so!
PS - same distance - 50cm or 1m??
I did some similar experiments a few days ago with both a pair of Peerless XXLS 12” and a pair of SB Acoustics SB17NBAC and got pretty much the same results as bushmeister. No screenshots because I thought that the 6 dB increase was a given. It actually seemed to work pretty high up in frequency. Purely anecdotal but I’m not letting go of the concept just yet…
@ayebee: what is "pretty high in frequency" in your case?
That's one unanswered question for me, I guess: what will happen at higher frequencies and off axis? I doubt there are any sets of measurements at lots of off axis angles. Are there any models that would show the off axis behavior at higher frequencies?
Anyway, this is all very interesting. I guess I had written off these kind of configurations and was confusing them with the "second order gradient" type systems, which have more low frequency losses compared to a dipole (a first order gradient).
That's one unanswered question for me, I guess: what will happen at higher frequencies and off axis? I doubt there are any sets of measurements at lots of off axis angles. Are there any models that would show the off axis behavior at higher frequencies?
Anyway, this is all very interesting. I guess I had written off these kind of configurations and was confusing them with the "second order gradient" type systems, which have more low frequency losses compared to a dipole (a first order gradient).
Ah too late. That's OK, nevermind, there is no real motivation to re-measure at this point. You could try it next time. It won't make any difference at low frequencies. I find it easier to do, since you can basically just lay the mic on the floor.
I was mainly testing up to 600 Hz, but even without a LP filter I couldn’t detect much in terms of anomalies due to phase issues higher than that. But again, didn’t save any measurements… It was mainly to test the thesis that clamshells were for subwoofer-frequencies only. I knew that the Legacy Whisper used a shallow LP-slope and a crossover point around 250 so it seemed somewhat plausible that the method might not introduce that many problems.
That could well explain it indeed. The measurement setup I used is shown in the picture. For the compound dipole I just added the 2nd woofer facing the first, so its "acoustic center" is quite more away from the mic.
I am quite happy with this outcome, as it allows for more compact building and hanging the compound woofer is much easier than single units! I do not have a direct use for it now, but knowing it can be done may inspire some future action.
Happy listening!
Erik
Attachments
It's folly to measure a dipole bass system so close. I do my dipole measurements at two or three meters indoors or five meters outdoors. Mic always on the ground to eliminate floor bounce. LP measurements are always good to add to the mix too.
It does appear there may be an increase in sensitivity. I wonder if this translates into a 6dB increase in maximum output, or if max o/p is essentally the same as a single driver.
It does appear there may be an increase in sensitivity. I wonder if this translates into a 6dB increase in maximum output, or if max o/p is essentally the same as a single driver.
Outdoor measurements are best of course, but not always possible. Distance to mic can be surprisingly short with no-baffle woofers, I have tested that too naturally. Room modes and first reflections cant be avoided indoors with long gating, but they are consistent in measurement results (off-axis by rotating source and trying different eq and xo slopes), so they can be easily eliminated from analysis.
I can't now do this kind of tests myself, because I don't have woofers available. Still after reading all previous, I can't see compound/isobaric worth the investment, unlike side by side or W frame like LX521, R16 or NaO note.
Prototyping a Dipole Bass System
H-frame or W-frame OB bass?
Gainphile: R16 - Constant Directivity Loudspeakers
I can't now do this kind of tests myself, because I don't have woofers available. Still after reading all previous, I can't see compound/isobaric worth the investment, unlike side by side or W frame like LX521, R16 or NaO note.
Prototyping a Dipole Bass System
H-frame or W-frame OB bass?
Gainphile: R16 - Constant Directivity Loudspeakers
Last edited:
Given SL's mathematical modelling of the compound/clamshell configuration - as seen in the celestion 6000 and legacy audio speakers - he has shown there is no difference between placing one behind the other or having them side by side.
His modelling predicts the same summation at low frequencies - and this is born out by multiple experimental evidence - from both myself and others.
If we therefore assume his modelling and our experiments are valid (which I believe they are) and that there is no difference in where we place the two woofers - as long as they are in close proximity to each other compared to the wavelengths they are producing (usually 4 meters and longer) and in phase with each other - then I don't see how there couldn't be a concurrent increase in Max SpL - we are after all - doubling the VD, power handling, etc. etc. - we are doubling the sub-woofers!
I would need a fairly robust reason for this not to be true - and experimental evidence to back it up.
I do not see any voodoo or black magic at work here. We are simply doubling our subwoofer count with the expected increase in output.
Whether this arrangement is worth it, given we can likely achieve better results with a single woofer in an appropriate H-baffle is an entirely different matter.
As I have previously explained - for me, having two 30" H baffles in my mancave was not an option, and the clamshell arrangement I am using is incredibly convenient for both the suspension of them - avoiding all mechanical transmission of energy into the open baffle frame - and the space saving of this arrangement.
If I were using very expensive 15" or 18" subwoofers this might be a different call and I would likely look into other options.
Personally I never doubted SL's modelling, it is both intuitive and seems completely logical given the very low frequencies in question, I feel I can put this matter to rest - but of course I would be very happy to be proven wrong if experimental evidence shows otherwise!
Just saying something from my intuition. There is a consensus that in dipole bass the drivers must move a lot of air. A driver moves air in phase and opposite phase, and when the two flows meet, they cancel each other, forming the figure of 8. This can be fought by increasing the distance before the two "airs" meet. Having two drivers side by side doubles the amount of moved air because of double moving surface, so for me it is intuitive that one gets the 6dB SPL increase with same individual drivers excursion. Now, if the drivers are mounted in the compound dipole, my initial thinking is that there is no increase in the amount of moved air, because the air around the second driver is already put in motion by the first driver. Still, it appears that there is still the 6dB increase, so there is more to SPL than moving air?!
I read some posts that commented that the compound woofer would have a different dipole pattern that a single woofer, actually beaming more: more output on axis and "bigger" nulls to the side. There is a lot of that info, and an interesting white paper from the Celestion 6000 system in this thread https://www.diyaudio.com/forums/multi-way/46111-legacy-whisper-bass.html#post3897099
It is sad that it is so hard to do a decent measurement of bass speakers. We live in an appartment, 4th flour, and I already thought about making a construction that would allow me to hang the speakers out of the balcony, having the mic (red stripe) also out of the balcony. This should avoid a lot of reflections... OTOH, ground plane outside should also tell a lot about the figure of 8 characteristic.
I read some posts that commented that the compound woofer would have a different dipole pattern that a single woofer, actually beaming more: more output on axis and "bigger" nulls to the side. There is a lot of that info, and an interesting white paper from the Celestion 6000 system in this thread https://www.diyaudio.com/forums/multi-way/46111-legacy-whisper-bass.html#post3897099
It is sad that it is so hard to do a decent measurement of bass speakers. We live in an appartment, 4th flour, and I already thought about making a construction that would allow me to hang the speakers out of the balcony, having the mic (red stripe) also out of the balcony. This should avoid a lot of reflections... OTOH, ground plane outside should also tell a lot about the figure of 8 characteristic.
Attachments
My reasoning goes, if two radiating membranes are close to each other and vibrating in same phase, the pressure variation between them is zero (when wavelength is longer than gap) . The summed radiated power will be same as a single radiating cone, and effective volume displacement and cone area are the same. Only benefit will be better motor control of cone movement and perhaps less distortion. Max spl will be the same because Xmax is the limiting factor. Bi-amping, parallel vs serial electric connection will make a difference in spl per voltage (sensitivity of the system).
I am illiterate about mathematics, but this is the conlusion I got years ago when reading Linkwitz's and Kreskowsky's pages. And remember, SL or JK never used this clamshell config himself.
Perhaps @john k... will contribute?
Isobaric
NaO Note II RS
Tech
I am illiterate about mathematics, but this is the conlusion I got years ago when reading Linkwitz's and Kreskowsky's pages. And remember, SL or JK never used this clamshell config himself.
Perhaps @john k... will contribute?
Isobaric
NaO Note II RS
Tech
Last edited:
My hunch is that there is an increase in sensitivity due to coupling but no improvement in max SPL as the max swept volume will still be the same as for a single driver.
If using active crossovers, the sensitivity of the woofer is not important so I see no value in this configuration.
This configuration is useful only on a passive crossover design which requires higher woofer sensitivity.
p.s. as the side-by-side configuration will require lower excursion for the same SPL, distortion should also be lower compared with the coupled configuration.
If using active crossovers, the sensitivity of the woofer is not important so I see no value in this configuration.
This configuration is useful only on a passive crossover design which requires higher woofer sensitivity.
p.s. as the side-by-side configuration will require lower excursion for the same SPL, distortion should also be lower compared with the coupled configuration.
"My reasoning goes, if two radiating membranes are close to each other and vibrating in same phase, the pressure variation between them is zero (when wavelength is longer than gap) . The summed radiated power will be same as a single radiating cone."
I do get this viewpoint and understand the reasoning behind this, but this is not correct - as confirmed by multiple measurements and SL's modelling.
I think if the space between the drivers were enclosed (i.e. the two subwoofer faces connected by a sealed tube) it would behave more as you suggest.
I do get this viewpoint and understand the reasoning behind this, but this is not correct - as confirmed by multiple measurements and SL's modelling.
I think if the space between the drivers were enclosed (i.e. the two subwoofer faces connected by a sealed tube) it would behave more as you suggest.