What cocolino and his umbrella analogy seems to be suggesting is the air in the space between enclosures is being dragged in the direction of the arrow so that the space between the drivers is providing additional surface area giving Sd+Sd+more.
Actually, when the cones initially move forward, the air in between the enclosures moves against the arrow.
Of course if the baffle is contiguous, the analogy breaks down altogether.
Actually, when the cones initially move forward, the air in between the enclosures moves against the arrow.
Of course if the baffle is contiguous, the analogy breaks down altogether.
Attachments
from "Loudspeaker and Headphones Handbook",
author John Borwick:
" The acoustic impedance acting on a single source is changed by other sources being in close proximity. ....So, at low frequencies, the radiation resistance of each source is doubled, and the power output of each, assuming that U remains constant in spite of the increased loading, is also doubled. The total power output from both sources, as compared with a single source, is quadrupled...."
In practice, Genelec says:
author John Borwick:
" The acoustic impedance acting on a single source is changed by other sources being in close proximity. ....So, at low frequencies, the radiation resistance of each source is doubled, and the power output of each, assuming that U remains constant in spite of the increased loading, is also doubled. The total power output from both sources, as compared with a single source, is quadrupled...."
In practice, Genelec says:
Attachments
Quoted from the Madisound board:
-----------------------------------------------------------------------------------------------------------
I can't believe that series/parallel driver thread below. Once and for all...
Posted By: John k... (34.216.252.64.snet.net)
Date: 8/1 2:03p.m.
(JPK) Forget this doubling cone area **** and all that. It has nothing to do with it. It's really simple. Take a driver, any driver, with a voltage of V= 1.0 applied across it. With V applied it generates a sound pressure of amplitude 1.0 (not using dB here, yet). Now take a second driver that also produces a sound pressure of 1.0 with V volts applied across it. Now, assume that these two drivers are operating in phase, each with 1.0 volts across it. Then the summed sound pressure is 1 + 1 = 2.0. Now last I knew a sound pressure 2.0 was 6 dB greated than a sound pressure of 1.0
Now, let's assume these two drivers also have the same impedance. If we wire them in parallel they will indeed each see 1.0 volts applied to both of them and the resulting SPL will be + 6dB. Note there is nothing here that says anything about cone area or what ever. All that can be said is that the output is 6dB greater. It can also be said, since the drivers have the same impedance, that there is twice the power being delivered into the system than there was for a single driver. The conclusion then is that since there is a 6dB increase in SPL with a 3dB increase in power, that the 2 driver system has higher efficiency than the single driver system. Since the additional SPL amounts to 3dB, from this we conclude that there is a 3dB SPL gain due to the effciency increase and since dB goes like 10 Log(efficiency), the efficiency of the two driver system has doubled. NOTE: this has nothing to do with doubling cone area. It's just that the two drivers are operaing in phase, just like an LR crossover between a 1" tweeter and a 8" woofer.
Now what happens if these two different drivers, which happen to have the same impedance are connected in series. Then each driver see 1/2 volt across it and the output sound pressure of each driver will drop to 1/2 as well. So the power into each driver is down 6dB and each driver's output is also down by 6 dB. But the summed output is 1/2 + 1/2 = 1.0 or 0dB, the same as each individual driver when operating by itself with 1 volt across it. Looking a little deeper, since the total power delivered to both drivers in the series configuration is -3dB but the summed output remained at 0dB, again showing that the two driver system has a 3dB gain due to increased efficiency.
But you can't just argue based on cone area or sensitivity. Two drivers, of any size, of any sensitivity, operating in phase at the same frequency and same SPL produce a 6dB increase in SPL over the single driver. It's just the vector sum of the sound pressure of each driver. The efficiency increase depends on the total power input and is only a factor of two when the drivers have the same base efficiency. For example, suppose driver A produces a sound pressure of 1 with 1 watt input and driver B produces a sound pressure of 1 with 1/2 watt input. Then, holding the power into each driver constant, we get a total power input of 1.5 watts with an output sound pressure of 2.0, or +6dB. But there is only a 1.76 dB increase in total power relative to 1 watt, so in this case there is a gain in SPL due to efficieny increases of 4.23 dB or the efficiency of this 2 driver system increases by 2.654 relative to the less sensitive driver. Relative to the more sensitive driver the input power went up by a factor of 3 or or 4.777dB meaning the SPL gain from efficiency increase is only 1.228dB or an efficiency increase of 1.327.
I think the major source of confusion comes in because people look at the definition of efficiency and see that it depends on cone area squared. Thus the conclusion is that if the cone area is doubled than the SPL goes up by 6 dB, all other things constant. When you start talking about multiple drivers you can't go back to the singler driver efficiency relationship. It's not about that. It's about vector sums of the driver's sound pressure and the division of power between the drivers. With the same drivers in series or parallel the power division thing is aways +0dB in parallel per driver, +3dB total power, -6dB in series per driver, -3dB total power, and the vector sum is always +6dB over the single operating at the same power level (i.e. the same power level for parallel or -6dB for series).
Mvg Johan
-----------------------------------------------------------------------------------------------------------
I can't believe that series/parallel driver thread below. Once and for all...
Posted By: John k... (34.216.252.64.snet.net)
Date: 8/1 2:03p.m.
(JPK) Forget this doubling cone area **** and all that. It has nothing to do with it. It's really simple. Take a driver, any driver, with a voltage of V= 1.0 applied across it. With V applied it generates a sound pressure of amplitude 1.0 (not using dB here, yet). Now take a second driver that also produces a sound pressure of 1.0 with V volts applied across it. Now, assume that these two drivers are operating in phase, each with 1.0 volts across it. Then the summed sound pressure is 1 + 1 = 2.0. Now last I knew a sound pressure 2.0 was 6 dB greated than a sound pressure of 1.0
Now, let's assume these two drivers also have the same impedance. If we wire them in parallel they will indeed each see 1.0 volts applied to both of them and the resulting SPL will be + 6dB. Note there is nothing here that says anything about cone area or what ever. All that can be said is that the output is 6dB greater. It can also be said, since the drivers have the same impedance, that there is twice the power being delivered into the system than there was for a single driver. The conclusion then is that since there is a 6dB increase in SPL with a 3dB increase in power, that the 2 driver system has higher efficiency than the single driver system. Since the additional SPL amounts to 3dB, from this we conclude that there is a 3dB SPL gain due to the effciency increase and since dB goes like 10 Log(efficiency), the efficiency of the two driver system has doubled. NOTE: this has nothing to do with doubling cone area. It's just that the two drivers are operaing in phase, just like an LR crossover between a 1" tweeter and a 8" woofer.
Now what happens if these two different drivers, which happen to have the same impedance are connected in series. Then each driver see 1/2 volt across it and the output sound pressure of each driver will drop to 1/2 as well. So the power into each driver is down 6dB and each driver's output is also down by 6 dB. But the summed output is 1/2 + 1/2 = 1.0 or 0dB, the same as each individual driver when operating by itself with 1 volt across it. Looking a little deeper, since the total power delivered to both drivers in the series configuration is -3dB but the summed output remained at 0dB, again showing that the two driver system has a 3dB gain due to increased efficiency.
But you can't just argue based on cone area or sensitivity. Two drivers, of any size, of any sensitivity, operating in phase at the same frequency and same SPL produce a 6dB increase in SPL over the single driver. It's just the vector sum of the sound pressure of each driver. The efficiency increase depends on the total power input and is only a factor of two when the drivers have the same base efficiency. For example, suppose driver A produces a sound pressure of 1 with 1 watt input and driver B produces a sound pressure of 1 with 1/2 watt input. Then, holding the power into each driver constant, we get a total power input of 1.5 watts with an output sound pressure of 2.0, or +6dB. But there is only a 1.76 dB increase in total power relative to 1 watt, so in this case there is a gain in SPL due to efficieny increases of 4.23 dB or the efficiency of this 2 driver system increases by 2.654 relative to the less sensitive driver. Relative to the more sensitive driver the input power went up by a factor of 3 or or 4.777dB meaning the SPL gain from efficiency increase is only 1.228dB or an efficiency increase of 1.327.
I think the major source of confusion comes in because people look at the definition of efficiency and see that it depends on cone area squared. Thus the conclusion is that if the cone area is doubled than the SPL goes up by 6 dB, all other things constant. When you start talking about multiple drivers you can't go back to the singler driver efficiency relationship. It's not about that. It's about vector sums of the driver's sound pressure and the division of power between the drivers. With the same drivers in series or parallel the power division thing is aways +0dB in parallel per driver, +3dB total power, -6dB in series per driver, -3dB total power, and the vector sum is always +6dB over the single operating at the same power level (i.e. the same power level for parallel or -6dB for series).
Mvg Johan
Actually, it does- it's implicit in his argument. If you double the number of drivers putting out 1.0, you've doubled the cone area (you've doubled the motor, too). He's dead right otherwise, especially in his warning about confusion between efficiency and sensitivity, which is the source of 3dB of the "gain."
About your 1.0+1.0 = 2.0 and that signifying a 6db increase.... its still only 3db. Double the power, and SPL goes up by 3db.
I was messing around with winISD to see what SPL it predicts for two drivers versus one. They show 6db(at any frequency, I guess they dont take into account distance between drivers). But when I played with the parameter box and doubled or halved Sd the results were exactly the opposite. Double in Sd caused a 6db DECREASE in maxSPL and halving Sd caused a 6db INCREASE in SPL.
Now I think that is kinda flawed due to the fact that changing Sd will change nearly every other parameter. The transfer function, however, doesnt change with a different Sd imput, but the max SPL does....
I also did a sim to see how many drivers it would take(mutually coupled) to achieve over 120db @1w 1m
3000 Tumults apparently....
WinISD also does some funky stuff when you use closed box and try to simulate infinite baffle. As the box keeps getting bigger and bigger, SPL and transfer function begin to decrease
I was messing around with winISD to see what SPL it predicts for two drivers versus one. They show 6db(at any frequency, I guess they dont take into account distance between drivers). But when I played with the parameter box and doubled or halved Sd the results were exactly the opposite. Double in Sd caused a 6db DECREASE in maxSPL and halving Sd caused a 6db INCREASE in SPL.
Now I think that is kinda flawed due to the fact that changing Sd will change nearly every other parameter. The transfer function, however, doesnt change with a different Sd imput, but the max SPL does....
I also did a sim to see how many drivers it would take(mutually coupled) to achieve over 120db @1w 1m
3000 Tumults apparently....
WinISD also does some funky stuff when you use closed box and try to simulate infinite baffle. As the box keeps getting bigger and bigger, SPL and transfer function begin to decrease
If you double Sd then you have to double the motor also.... hmmm
well this does not seem so true when you look at adire's brahma series.
All of the motors are the exact same(or at least they seem to be judging by the exact same Re BL Xmax and Le) but cone sizes are different.
Brahma 10: Sd-300cm^2 Sensitivity-82.3db @ 1w
Brahma 12: Sd-436cm^2 Sensitivity-84.4db @ 1w
Brahma 15: Sd-749cm^2 Sensitivity-87.6db @ 1w
Now the Q's are pretty close to one another so I dont think that is what is changing the sensitivity.... but Vas is dramtically different.
It seems to me that the Vas is what is really making the difference here.
So I did some sims in winISD and this time instead of playing with Sd I changed Vas..... The conclusion, double Vas= +6dB
What does GIGO mean?
well this does not seem so true when you look at adire's brahma series.
All of the motors are the exact same(or at least they seem to be judging by the exact same Re BL Xmax and Le) but cone sizes are different.
Brahma 10: Sd-300cm^2 Sensitivity-82.3db @ 1w
Brahma 12: Sd-436cm^2 Sensitivity-84.4db @ 1w
Brahma 15: Sd-749cm^2 Sensitivity-87.6db @ 1w
Now the Q's are pretty close to one another so I dont think that is what is changing the sensitivity.... but Vas is dramtically different.
It seems to me that the Vas is what is really making the difference here.
So I did some sims in winISD and this time instead of playing with Sd I changed Vas..... The conclusion, double Vas= +6dB
What does GIGO mean?
Remember the issue at hand- doubling up drivers. Double the drivers, double the Sd, but you also have doubled the motors- you're not using a single coil to drive both diaphragms.
GIGO = "Garbage in, garbage out." It's a common saying in the computer world- a calculation is only as good as the assumptions made.
GIGO = "Garbage in, garbage out." It's a common saying in the computer world- a calculation is only as good as the assumptions made.
It's not complicated. You've got two acoustic sources. How do they add? (Answer: vectorially)
A separate question is, what happens when you put two of the same speakers in the same box. It's a valid question, it's just not what I was talking about (nor Bill nor John K). If you stick two drivers in the same box, you can use T-S theory to tell you what the net near-field frequency response is for each driver. Then you can use the 6dB rule we're discussing here (and everyone agrees that it's 6dB experimentally) to predict the sensitivity in the frequency range where the drivers are acoustically close.
A separate question is, what happens when you put two of the same speakers in the same box. It's a valid question, it's just not what I was talking about (nor Bill nor John K). If you stick two drivers in the same box, you can use T-S theory to tell you what the net near-field frequency response is for each driver. Then you can use the 6dB rule we're discussing here (and everyone agrees that it's 6dB experimentally) to predict the sensitivity in the frequency range where the drivers are acoustically close.
Alright so I tried this.....
On winISD I simulated a closed box with two adire tumults with a Q of .707
Then I simulated a single Tumult in the same sized box. So I figure if the mutual coupling makes two drivers act as one with double the radiating surface area, double motor strength, and double vas, then to simulate that I doubled VAS motor strength and surface area.
The result was the same max SPL graph. But isnt the power handling also doubled? When I double that, the single driver equivilant becomes 3db louder on the max output.
What is wrong here?
On winISD I simulated a closed box with two adire tumults with a Q of .707
Then I simulated a single Tumult in the same sized box. So I figure if the mutual coupling makes two drivers act as one with double the radiating surface area, double motor strength, and double vas, then to simulate that I doubled VAS motor strength and surface area.
The result was the same max SPL graph. But isnt the power handling also doubled? When I double that, the single driver equivilant becomes 3db louder on the max output.
What is wrong here?