Hi,
As I recall you stated you compared a free air driver compared to the compound arrangement, details are
lacking, but it sounds like you have not compared a H frame single driver to the equivalent compound pair.
rgds, sreten.
Last edited:
john k...
I can follow your explanation, but where is the point where two dipoles become one isobaric system?
I'm under the impression that cuibonos bass dipole still is something like this:
From the looks of it I could consider it as a combination of two very short U frame dipoles separated by a very short distance (compared to the wavelength involved). But isn't it already a "leaky" tube with one driver at each end? Or a "leaky" H frame? What if I built a true H frame with the same depth as his construction, but only two drivers instead of four?
Rudolf
I can follow your explanation, but where is the point where two dipoles become one isobaric system?
I'm under the impression that cuibonos bass dipole still is something like this:
An externally hosted image should be here but it was not working when we last tested it.
From the looks of it I could consider it as a combination of two very short U frame dipoles separated by a very short distance (compared to the wavelength involved). But isn't it already a "leaky" tube with one driver at each end? Or a "leaky" H frame? What if I built a true H frame with the same depth as his construction, but only two drivers instead of four?
Rudolf
>But isn't it already a "leaky" tube with one driver at each end? Or a "leaky" H frame?
Yes.
Another obvious difference compared to a single flipped pair is here the set of cavities all have lower volume, maybe important at higher turnover points. At resonant frequency with a pure sine drive there should be no practical difference with or without side covers, and it's still impossible to get higher SPL than the driver excursion allows for a pair each mounted in their own baffle hole in an equivalent dimensioned H frame. Cone stress due to acoustic load is roughly halved in the double configuration, minus the air mass in the middle, loss due to shear, turbulence..
Yes.
Another obvious difference compared to a single flipped pair is here the set of cavities all have lower volume, maybe important at higher turnover points. At resonant frequency with a pure sine drive there should be no practical difference with or without side covers, and it's still impossible to get higher SPL than the driver excursion allows for a pair each mounted in their own baffle hole in an equivalent dimensioned H frame. Cone stress due to acoustic load is roughly halved in the double configuration, minus the air mass in the middle, loss due to shear, turbulence..
Last edited:
Rudolph - yes, that is one of my bass arrangements. But the one I am currently working on has a sealed box/chamber, with two drivers on either side, connected in parallel, working as a dipole (+front, -back). I guess I'll have to post photos sometime, but the graphs I posted earlier are of the closed box version, which I thought was the topic of the OPs question.
Like I asked, has anyone actually built and measured these variations, because I have. I wouldn't let theory stand in the way of actual results.
I might add that I first started down this line of thought while dealing with MJK's Alpha15 H-frame. I added a second driver to the H-frame, just mounting it face to face of the other, wire in in parallel, but out of phase. The results were substantially worse than just a single driver. I also tried it with the GW-210, with the same results. From there I just spaced two drivers front to back, without any baffle or chamber, wired in parallel and in phase. In the bass, it doubled the total SPL output, compared to a single driver. Really, finding an 'appropriate' space is just empirical - but for me its been about 5" between cones, with the drivers facing the same direction.
Like I asked, has anyone actually built and measured these variations, because I have. I wouldn't let theory stand in the way of actual results.
I might add that I first started down this line of thought while dealing with MJK's Alpha15 H-frame. I added a second driver to the H-frame, just mounting it face to face of the other, wire in in parallel, but out of phase. The results were substantially worse than just a single driver. I also tried it with the GW-210, with the same results. From there I just spaced two drivers front to back, without any baffle or chamber, wired in parallel and in phase. In the bass, it doubled the total SPL output, compared to a single driver. Really, finding an 'appropriate' space is just empirical - but for me its been about 5" between cones, with the drivers facing the same direction.
Cuibono, I didn't mean to imply that you were joking or worse when you said you got greater output with the doubled/spaced arrangement. For drivers with very low BL like the ones you mentioned the output might increase just because there is less electrical loss. Each cone only gets half the acoustic load, while there are two motors handling it. The Q should remain the same because the resonant system is just doubled in size, otherwise unaltered. There is a possibility that the increase in bass performance is just because of the H-frame effect of the space, increasing baffle performance for lower frequencies. You seem to be in as good a position as anyone to investigate this further. I'm not arguing either way, just trying to make sense of it.
When I said "impossible to get higher SPL than the driver excursion allows for a pair" I didn't mean that sensitivity or efficiency wouldn't change, just that the excursion limit is still the excursion limit.
When I said "impossible to get higher SPL than the driver excursion allows for a pair" I didn't mean that sensitivity or efficiency wouldn't change, just that the excursion limit is still the excursion limit.
I don't think it is really worth discussing. There is no isobaric action here. The term isobaric (constant pressure) is really misleading in the first place. A true isobaric system with front driver coupled to the rear driver through a sealed chamber.
Here is a picture from my web page on isobaric systems:
An externally hosted image should be here but it was not working when we last tested it.
K1 is the compliance of the rear box of an isobaric system. K2 is the compliance of the chamber coupling the two, X1 and X2 are the displacements of the drivers. For there to be any isobaric action x1 -x2 must vary with time since the coupling force is k2 * (x1-x2). x1-x2 becomes different because of the force on the rear driver, x1*k1 - (x1-x2)*k2 is different than the force on the front driver, k2 * (x1-x2). If the rear box is not present then k1 = 0 and the force on the driver is only k2 * (x1-x2) which will be zero unless x1 and x2 are different. But if the drivers are identical and driven by the same signal (with correct polarity), then x1 = x2 and the force exerted between the drivers is zero regardless of whether the chamber is sealed or open.
If you look at the response of the woofer system in your picture you will surely be able to generate the same low frequency response with only two drivers by designing an H frame with the correct length.
"same low frequency response"
Only if the single pair replacing the double pair has the same response as the double pair.
You might not want to ignore that you've doubled the motor strength and cone strength while the acoustic load remained constant. Regarding the theoretical benefits of that, the double OB might be more isobaric than isobaric, since the cone facing the chamber in the box gets loaded a lot harder than the front cone and therefore they're never moving the same even if "clamshelled". But resonant response doesn't get altered in the OB arrangement with any number of stacked drivers because there is no K1, like you said, just a number of hopefully identical resonant systems running in unison.
Only if the single pair replacing the double pair has the same response as the double pair.
You might not want to ignore that you've doubled the motor strength and cone strength while the acoustic load remained constant. Regarding the theoretical benefits of that, the double OB might be more isobaric than isobaric, since the cone facing the chamber in the box gets loaded a lot harder than the front cone and therefore they're never moving the same even if "clamshelled". But resonant response doesn't get altered in the OB arrangement with any number of stacked drivers because there is no K1, like you said, just a number of hopefully identical resonant systems running in unison.
Last edited:
Hello again,
besides any "efficiency" discussion, where could
be a benefit in splitting an H-Frame using 2 drivers ?
Given a rather deep and tube shaped undamped
H-Frame is split like
1-driver-2-driver-1 in depth,
the wavelengh for resonance of both
single end closed tubes will be 4*1=4.
The double sided closed tube in the middle
will have the same lowest resonant wavelength
with 2*2=4.
The undamped single driver H-Frame of same
depth
2-driver-2
will have resonant wavelength 4*2=8.
Which would yield an upper non resonant frequency
limit for the split one being one octave higher.
One octave increase in non resonant bandwidth for
the undamped case could be seen as an advantage
i think. Of cause for the correct estimation
of "optimum" split ratios, aperture/shape correction
will have to be applied.
As soon as the "inner" distance of the drivers
compares to some significant fraction of wavelength,
there will be unwanted radiation from the "inner" dipole,
as has been already pointed out by John K, given the
inner "chamber" is open at its circumference ...
So if someone wants to make use of the higher
"non resonant bandwitdh", proposal would be to have
the connecting duct closed (and maybe also suffciently
absorbent). It could be made of felt layers e.g. since
it being perfectly airtight seems not necessary, when
referring to John K:
"k2 * (x1-x2) which will be zero ..."
Just my 2ct ...
Last edited:
I have not thought about the inner tube being excited
at both ends ... sorry.
We have to rethink that 🙄
at both ends ... sorry.
We have to rethink that 🙄
...
but since both ends of the tube are exited in antiphase
the resonances for
k * lambda/2 should be maintained (gained!) for k being odd
and canceled for k being even.
Please check someone ...
but since both ends of the tube are exited in antiphase
the resonances for
k * lambda/2 should be maintained (gained!) for k being odd
and canceled for k being even.
Please check someone ...
I knew someone would bring that up. Yes, they are done with a short window. I don't know why I kept these measurements, as I almost always use longer, more appropriate windowing. But you'll have to take it on faith, I guess. Only people familiar with SoundEasy (the measurement program I use) will know that changing the windowing doesn't change the overall trend of the SPL response, it only makes it look like it has more fine detail. Maybe I'll post more measurements, as I'm preparing to do another set on them.
Yes, your right, this isn't an isobaric system, nor is one with a sealed chamber. But that is just a misused word, and is not particularly important to the conversation. The OP asked if there are any benefits of using two drivers, in dipole configurations, with a chamber separating them.
From a theory point of view, you are saying there isn't. I'm saying I've made several bass systems where I've measured a useful improvement. If people want, I'll start a new thread where I detail construction and measurement.
Sorry if this back and forth has gotten tiring to anyone.
Last edited:
Hopefully no one is actually getting tired already. I had trouble seeing it at first but I think the thread has essentially already touched on all the performance changes/benefits. By now I'd only point out that excluding higher frequency effects the same result could mainly be achieved by an appropriate choice of different higher quality drivers having same Q but heavier cone and bigger magnet with a deeper baffle. If you must have the cone characterics of the lighter drivers for some reason there would be no way to match the double setup. That could be a very serious advantage. There could be any number of lesser important effects going on but I wouldn't expect to discover anything major.
Last edited:
Okay, long reply.
Here are some measurements, hot of the press. I tested four setups. All tests were done 9 feet off the ground, mic distance of 60" from the approximated acoustic centers, gating was 25ms, and the driving signal was kept constant.
First, a single GW210/8 driver:
Second, two drivers with a 4.75" space, wired in parallel and in phase:
Third, two drivers wired the same but with a 5.25" chamber between them:
Fourth, a single driver with a 2.5" wide foam-core circular baffle:
Here is another shot of the overall test setup:
Here are some measurements, hot of the press. I tested four setups. All tests were done 9 feet off the ground, mic distance of 60" from the approximated acoustic centers, gating was 25ms, and the driving signal was kept constant.
First, a single GW210/8 driver:
An externally hosted image should be here but it was not working when we last tested it.
Second, two drivers with a 4.75" space, wired in parallel and in phase:
An externally hosted image should be here but it was not working when we last tested it.
Third, two drivers wired the same but with a 5.25" chamber between them:
An externally hosted image should be here but it was not working when we last tested it.
Fourth, a single driver with a 2.5" wide foam-core circular baffle:
An externally hosted image should be here but it was not working when we last tested it.
Here is another shot of the overall test setup:
An externally hosted image should be here but it was not working when we last tested it.
Here are the raw results:
This first set is of the single, dual, and chambered setups, from bottom to top (in the bass):
Here is just the single driver compared to the chambered setup (single is on bottom):
Here is the single driver compared to the baffled driver:
Here is the single, baffled, and chambered setup (from bottom to top):
The baffled driver should have the same dipole moment as the chambered setup. Notice the chambered setup is 3dB hotter than the baffled driver.
Notice that dual naked drivers gives a 6dB increase in output compared to a single naked driver (as would be expected). This can be extrapolated to my current woofer arrangement (which Rudolf posted earlier), which is similar, but with baffles. The bottom line is that for drivers on axis with each other, and open in between them, they act as independent sources, and give 6dB gain for each doubling, in the bass region, where the size of the baffle is small compared to the wavelengths involved.
I'm working on a new woofer setup, and may post more measurements today showing more of the differences between a single driver and two on-axis drivers.
This first set is of the single, dual, and chambered setups, from bottom to top (in the bass):
An externally hosted image should be here but it was not working when we last tested it.
Here is just the single driver compared to the chambered setup (single is on bottom):
An externally hosted image should be here but it was not working when we last tested it.
Here is the single driver compared to the baffled driver:
An externally hosted image should be here but it was not working when we last tested it.
Here is the single, baffled, and chambered setup (from bottom to top):
An externally hosted image should be here but it was not working when we last tested it.
The baffled driver should have the same dipole moment as the chambered setup. Notice the chambered setup is 3dB hotter than the baffled driver.
Notice that dual naked drivers gives a 6dB increase in output compared to a single naked driver (as would be expected). This can be extrapolated to my current woofer arrangement (which Rudolf posted earlier), which is similar, but with baffles. The bottom line is that for drivers on axis with each other, and open in between them, they act as independent sources, and give 6dB gain for each doubling, in the bass region, where the size of the baffle is small compared to the wavelengths involved.
I'm working on a new woofer setup, and may post more measurements today showing more of the differences between a single driver and two on-axis drivers.
Help me now: how/why is it not relevant that in doubling the driver (wired in parallel, presumably,) you are delivering twice the power to the assembly with constant voltage applied? That would explain why the single baffled has 3 dB less output.
Is the point that coupling them with an enclosed tube yields more net SPL gain than not?
Is the point that coupling them with an enclosed tube yields more net SPL gain than not?
Last edited:
Here are some measurements of my GWx4 woofer system I did today, under the same conditions as above.
I was measuring the difference between just two drivers in the front section, compared to four drivers in two parallel planes, with the rear drivers on-axis with the front drivers. The front drivers are wired in parallel, out of phase, and for their measurement, I removed the rear two drivers.
I then added the rear drivers, and they were also wired in parallel, out of phase - BUT the two groups (front two and rear two) were not wired together. They were driven by separate amps, as I couldn't wire them in parallel (which would equal a 2 ohm load). I could have wired them all in series, but that would have made the comparison more complicated.
Here are the results, with the two driver test on the bottom, and the four driver test on top - again, the area of interest is between 40-200Hz:
Notice that at 40Hz, the four driver system is up 6dB. At 100Hz, it is up 4dB. The point is that multiple woofers on-axis work as independent sources, which they largely seem to as shown by these measurements.
An externally hosted image should be here but it was not working when we last tested it.
I was measuring the difference between just two drivers in the front section, compared to four drivers in two parallel planes, with the rear drivers on-axis with the front drivers. The front drivers are wired in parallel, out of phase, and for their measurement, I removed the rear two drivers.
I then added the rear drivers, and they were also wired in parallel, out of phase - BUT the two groups (front two and rear two) were not wired together. They were driven by separate amps, as I couldn't wire them in parallel (which would equal a 2 ohm load). I could have wired them all in series, but that would have made the comparison more complicated.
Here are the results, with the two driver test on the bottom, and the four driver test on top - again, the area of interest is between 40-200Hz:
An externally hosted image should be here but it was not working when we last tested it.
Notice that at 40Hz, the four driver system is up 6dB. At 100Hz, it is up 4dB. The point is that multiple woofers on-axis work as independent sources, which they largely seem to as shown by these measurements.
No. If the dipole separation of the chambered setup is 5.25", you have to add those 5.25" to the RADIUS of the baffle, not to the diameter. In your case the 3 dB difference of the baffled driver to either single driver and chambered setup is exactly in line with known theory.
Nice to see that. I wasn't quite sure if dipoles really would work independently at such short distances. Thanks for showing it.
And my point still is, that two drivers in a H frame of the same outer dimensions as your construction would give the same level as the four of yours. 😉
As Linkwitz and John K already mentioned: The distance between the drivers is largely irrelevant. You can always merge two into one. The only relevant factor is the distance separating the dipole sorces.
Rudolf
Exactly. It doesn't matter where the drives are or how they are arranged. Just start figuring path length differences and sum. Obviously with the last set up, if you reduce the separation between the front and rear sets then the rear radiation from the front set and the front radiation from the rear set just cancel. If S is the separation, as S goes to zero the response becomes that of one set.
All these comparisons are apples to oranges. Of course they all have different SPL. Why would you expos anything else?
- Status
- Not open for further replies.