Instead of one open baffle 18" or 21" bass driver I am investgating the possibility of 16 10" drivers in Ripole configuration.
What can I expect sensitivity wise? In theory if the drivers are wired to a 8ohm load the sensitivity will go up by 12db right? So if the drivers are 90db sensitive combined they are 102db sensitive. But I read somewhere it might be a bit less due to losses which seems reasonable. Otherwise you could build an infinitely sensitive array and blow up an entire city.
So let's say the sensitivity ends up being 99db. That means the same amp plays 9db louder with 1w@1m
But instead of one speaker rated at let's say 100w now I have 16 drivers. So intuitively I would think if I needed a 100w amp for one driver I would want a 1600w amp for 16 drivers.
I guess I want a higher wattage amp for 16 drivers but probably not 16x.
I don't know much about electronics so please be gentle with your explanations.
What can I expect sensitivity wise? In theory if the drivers are wired to a 8ohm load the sensitivity will go up by 12db right? So if the drivers are 90db sensitive combined they are 102db sensitive. But I read somewhere it might be a bit less due to losses which seems reasonable. Otherwise you could build an infinitely sensitive array and blow up an entire city.
So let's say the sensitivity ends up being 99db. That means the same amp plays 9db louder with 1w@1m
But instead of one speaker rated at let's say 100w now I have 16 drivers. So intuitively I would think if I needed a 100w amp for one driver I would want a 1600w amp for 16 drivers.
I guess I want a higher wattage amp for 16 drivers but probably not 16x.
I don't know much about electronics so please be gentle with your explanations.
sensitivity increase is the result of on-axis directivity increase. So your net efficiency does not increase, you only direct it to a much smaller area. Efficiency may increase a bit in the lower frequencies due to improved loading, resulting in a lower cutoff frequency.
No, its not just a matter of directivity. There's a real efficiency gain from the increased cone area. But you need to bear in mind that a driver's nominal sensitivity is pretty meaningless when it comes to open baffle bass - the final response will be determined just as much by the baffle size and the system Fs and Qts.
There is no free lunch: you do not get "free" sensitivity for long because you need to be able to deliver more and more current and the amplifier cannot do that.
Example:
1 Driver: 90dB/W at 8 Ohms
2 Drivers in parallel: 96dB/W, the load is 4 Ohms
4 Drivers in parallel: 102dB/W, the load is 2 Ohms
As you double the number of drivers the voltage sensitivity goes up by 6dB (actually even more, but let's keep it simple) but the load impedance is halved.
Taking this to the 16 driver level, the system in total would have +24dB voltage sensitivity but 1/16th of the impedance of a single driver.
Most amplifiers can handle a 4 Ohm load at low frequency, and a few (for home audio) can handle 2 Ohms. Below that the amplifier will struggle and probably go into protection soon at higher playback levels.
Example:
1 Driver: 90dB/W at 8 Ohms
2 Drivers in parallel: 96dB/W, the load is 4 Ohms
4 Drivers in parallel: 102dB/W, the load is 2 Ohms
As you double the number of drivers the voltage sensitivity goes up by 6dB (actually even more, but let's keep it simple) but the load impedance is halved.
Taking this to the 16 driver level, the system in total would have +24dB voltage sensitivity but 1/16th of the impedance of a single driver.
Most amplifiers can handle a 4 Ohm load at low frequency, and a few (for home audio) can handle 2 Ohms. Below that the amplifier will struggle and probably go into protection soon at higher playback levels.
The other point that has probably not been considered here is POWER. If you want the drivers to keep adding SPL you only get that if they are each drawing/dissipating the same amount of power.
So let's look at my example again, the driver that is 90dB/W with 8 Ohms nominal sensitivity. This is 1W to get 90dB from one driver.
With 2 in parallel, you must deliver 2 Watts to the system so that each one gets 1 W and their outputs add to +6dB.
The consequence is that, if you are paralleling N identical drivers you need N times that power compared to 1 driver when all 16 will be operated just as hard as the single driver.
Any good loudspeaker box modeling program should be able to show you some hard numbers for your driver in question.
Also, did you think about the sensitivity of the Ripole configuration compared to the MFG passband SPL spec? It will be much lower, maybe by 10dB or more depending on your LP filter or inductor value.
The ripole, without EQ, has a peak in the frequency response and below that the FR drops off steadily. You need a filter to flatten the response and this essentially cuts off or attenuates the higher sensitivity part of the response closer tot he peak, leaving a flat but low sensitivity passband. It's really not a good dipole setup except for very small rooms and close listening. Better would be a stack of U- or H-frames like Monte Kay built a long time ago:
http://www.mfk-projects.com/Home_Theatre/theatre_woofer.html
You could also do this with 10" drivers, but to get more SPL larger diameter drivers ARE better because of more Xmax*Sd.
So let's look at my example again, the driver that is 90dB/W with 8 Ohms nominal sensitivity. This is 1W to get 90dB from one driver.
With 2 in parallel, you must deliver 2 Watts to the system so that each one gets 1 W and their outputs add to +6dB.
The consequence is that, if you are paralleling N identical drivers you need N times that power compared to 1 driver when all 16 will be operated just as hard as the single driver.
Any good loudspeaker box modeling program should be able to show you some hard numbers for your driver in question.
Also, did you think about the sensitivity of the Ripole configuration compared to the MFG passband SPL spec? It will be much lower, maybe by 10dB or more depending on your LP filter or inductor value.
The ripole, without EQ, has a peak in the frequency response and below that the FR drops off steadily. You need a filter to flatten the response and this essentially cuts off or attenuates the higher sensitivity part of the response closer tot he peak, leaving a flat but low sensitivity passband. It's really not a good dipole setup except for very small rooms and close listening. Better would be a stack of U- or H-frames like Monte Kay built a long time ago:
http://www.mfk-projects.com/Home_Theatre/theatre_woofer.html
You could also do this with 10" drivers, but to get more SPL larger diameter drivers ARE better because of more Xmax*Sd.
Last edited:
Yes, but the OP was asking about wiring them to an 8 ohm load. I presume he intends to use nominal 8 ohm drivers, and would arrange them in 4 parallel branches each with 4 drivers in series. That would give a 12 dB gain and preserve the 8 ohm load.
We are talking low frequencies (<150hz). I wouldn't think directivity is a major factor here.
I am working on a high SPL low distortion 4 way open baffle. If I put a large subwoofer, bass and midrange on a baffle the tweeter will end up very high, substantially above ear level.
On a 60cm wide baffle I think I can squeeze an 8" mid and 2x10" bass drivers flanked on both sides by two 8 driver vertical Ripole arrays and then the tweeter ends up at a sane height while the Ripole driver area is equivalent to three 21 drivers. So excursion and this distortion can be kept low. And in Ripole there is force cancellation so little baffle excitement. And because the drivers are only 10" the Ripole cavity resonance is all the way up around 340hz.
But yes....lots of work.
You mentioned something like "90 dB 1W@1m", which would be efficiency (i.e., how efficient power is converted into SPL).
Sensitivity is about SPL for a given voltage.
Sorry, no, I don't believe I did.
In the case the OP asked about, multiple identical drivers wired so as to preserve the impedance of a single driver, there are gains in both efficiency and sensitivity, surely. I can't see a problem caused by a confusion between them here.
In the case the OP asked about, multiple identical drivers wired so as to preserve the impedance of a single driver, there are gains in both efficiency and sensitivity, surely. I can't see a problem caused by a confusion between them here.
Last edited: