Power tubes and heat-build up
- Thread starter Kit
- Start date
Help Support GroupDIY Audio Forum:
> I´ve would have thought that a closed box would have been more efficient. Any papers on this?
EV had a lightweight paper around 1977.
A cone makes sound on both sides. A sealed box traps the backwave. A ported box traps the backwave except in the half-octave around the box resonance. We are throwing away half the acoustic output.
This is only true if electro-acoustic efficiency is low. If the air-load were the main load on the motor, using the backwave would take more electric power for same overall efficiency. But 95+% of the load on a cone is the Mass, not the air-load. The mass does not change when we tinker the acoustic side. (Well, a teeny bit due to air mass, but you'd have to use a very small box to change the air mass enough to measure.)
So if the Thiele half-space reference efficiency is 2%, using both sides of the cone we have potential 4% electro-acoustic conversion efficiency. That is, if we can put that out-of-phase sound to good use.
In a typical performance situation, the back-wave bounces off the rear wall and joins the indirect sound that bounces around the room, increasing the reverberant field. Since the reverb is typically random phase, the out-of-phase rear-wave can join the party without cancellation.
Increasing the reverberant field is usually bad for clarity. However the presence of this rear-wave at the edge of the front-wave totally changes the radiation pattern. A small speaker in a closed box is an omni radiator. A small speaker on a baffle is a bi-directional radiator and gives a pattern like a simple ribbon microphone. Side radiation is suppressed, which reduces the reverberant field. The on-axis radiation rises a little. The net result is like the difference between two mikes, one omni and one bi-di: the bi-di "reaches" further through the reverberant field for directions along its axis, but performs poorly to the side. In many performance situations, you can aim the axis so the front lobe covers most of the crowd, and the falloff to the side tends to match the shorter distance to patrons on the side of the stage. This fails in very wide rooms, or outdoors (where the backwave just runs away), but is good more often than not.
The disadvantage of the open baffle is that the side of the face is a function of the lowest frequency you want. In free space, the side should be a half wavelength. For guitar, 84 inches or 7 feet! However the roll-off is much slower than box speakers, we can fudge. The classic Fender speakers and amps have a strong resonance at the bottom note, which can compensate the baffle loss. Say we aim for 160Hz, 42 inches on a side. Then if we put the baffle on the floor, we can cut half of it off: 42"x21". This is very nearly the size of the classic Fender Twin-12 cabinet: the Fender is a bit taller, less wide on the front, but has significant fold-back around the edges which increases the path length. The directivity advantages remain valid despite our cheats: there is no other way to get significant directivity at 80Hz in a package this small. A cone array or horn would have to be closer to 84 inches to project 80Hz as well as a Twin does. And 82" does not fit in the Econoline.
You can't fool mother nature. The baffle loss means the cone excursion rises fast from 160Hz to 80Hz. If you need a lot of POWER at 80Hz, you could do better with a box. Indeed guitarists working with a bass player don't normally ask for blow-down power at 80Hz; that's the bass's job. They use the 80Hz fundamental as harmonic foundation, not earthquake-shake.
EV had a lightweight paper around 1977.
A cone makes sound on both sides. A sealed box traps the backwave. A ported box traps the backwave except in the half-octave around the box resonance. We are throwing away half the acoustic output.
This is only true if electro-acoustic efficiency is low. If the air-load were the main load on the motor, using the backwave would take more electric power for same overall efficiency. But 95+% of the load on a cone is the Mass, not the air-load. The mass does not change when we tinker the acoustic side. (Well, a teeny bit due to air mass, but you'd have to use a very small box to change the air mass enough to measure.)
So if the Thiele half-space reference efficiency is 2%, using both sides of the cone we have potential 4% electro-acoustic conversion efficiency. That is, if we can put that out-of-phase sound to good use.
In a typical performance situation, the back-wave bounces off the rear wall and joins the indirect sound that bounces around the room, increasing the reverberant field. Since the reverb is typically random phase, the out-of-phase rear-wave can join the party without cancellation.
Increasing the reverberant field is usually bad for clarity. However the presence of this rear-wave at the edge of the front-wave totally changes the radiation pattern. A small speaker in a closed box is an omni radiator. A small speaker on a baffle is a bi-directional radiator and gives a pattern like a simple ribbon microphone. Side radiation is suppressed, which reduces the reverberant field. The on-axis radiation rises a little. The net result is like the difference between two mikes, one omni and one bi-di: the bi-di "reaches" further through the reverberant field for directions along its axis, but performs poorly to the side. In many performance situations, you can aim the axis so the front lobe covers most of the crowd, and the falloff to the side tends to match the shorter distance to patrons on the side of the stage. This fails in very wide rooms, or outdoors (where the backwave just runs away), but is good more often than not.
The disadvantage of the open baffle is that the side of the face is a function of the lowest frequency you want. In free space, the side should be a half wavelength. For guitar, 84 inches or 7 feet! However the roll-off is much slower than box speakers, we can fudge. The classic Fender speakers and amps have a strong resonance at the bottom note, which can compensate the baffle loss. Say we aim for 160Hz, 42 inches on a side. Then if we put the baffle on the floor, we can cut half of it off: 42"x21". This is very nearly the size of the classic Fender Twin-12 cabinet: the Fender is a bit taller, less wide on the front, but has significant fold-back around the edges which increases the path length. The directivity advantages remain valid despite our cheats: there is no other way to get significant directivity at 80Hz in a package this small. A cone array or horn would have to be closer to 84 inches to project 80Hz as well as a Twin does. And 82" does not fit in the Econoline.
You can't fool mother nature. The baffle loss means the cone excursion rises fast from 160Hz to 80Hz. If you need a lot of POWER at 80Hz, you could do better with a box. Indeed guitarists working with a bass player don't normally ask for blow-down power at 80Hz; that's the bass's job. They use the 80Hz fundamental as harmonic foundation, not earthquake-shake.
bcarso
Well-known member
That's a great discussion PRR---one of your classics!
On a related matter:
I'm collaborating with a friend on a highend home system that features a way-below-wavelength-wide open-back woofer, which crosses over to the mid-tweet panel at about 160Hz. The woofer drivers are these enormous things with double radial-neo motors from Aurasound, custom-made with suspensions tuned to the optimal stiffness for this application. The boost compensation for response falloff at lower frequencies entails huge amp powers at the lowest frequencies. OTOH he doesn't like to listen much at high SPLs, so it works for him at least.
The payoff is a much less critical room placement for uniform bass response. This guy doesn't care much for off-axis response though ("I want to listen to the music, not the room"), and the mid-tweet panels are highly directional. So it is a one-person-in-the-sweet-spot sort of system, at least for any kind of decent imaging, in that way not unlike the big flat electrostatic panels that beam like crazy.*
I'm mainly doing the electronics portion for fun, because of the near-lack of cost constraints, and the challenge of satisfying some of the nominal audiofiligree presumptions while delivering excellent measured performance. But who knows---there may be some people who will actually buy the things. Also, out of the exercise I've found a few circuit variants that can be applied to other things.
*It's not too surprising that this guy doesn't have a significant other in his life, despite his insistence on wanting one :razz: .
On a related matter:
I'm collaborating with a friend on a highend home system that features a way-below-wavelength-wide open-back woofer, which crosses over to the mid-tweet panel at about 160Hz. The woofer drivers are these enormous things with double radial-neo motors from Aurasound, custom-made with suspensions tuned to the optimal stiffness for this application. The boost compensation for response falloff at lower frequencies entails huge amp powers at the lowest frequencies. OTOH he doesn't like to listen much at high SPLs, so it works for him at least.
The payoff is a much less critical room placement for uniform bass response. This guy doesn't care much for off-axis response though ("I want to listen to the music, not the room"), and the mid-tweet panels are highly directional. So it is a one-person-in-the-sweet-spot sort of system, at least for any kind of decent imaging, in that way not unlike the big flat electrostatic panels that beam like crazy.*
I'm mainly doing the electronics portion for fun, because of the near-lack of cost constraints, and the challenge of satisfying some of the nominal audiofiligree presumptions while delivering excellent measured performance. But who knows---there may be some people who will actually buy the things. Also, out of the exercise I've found a few circuit variants that can be applied to other things.
*It's not too surprising that this guy doesn't have a significant other in his life, despite his insistence on wanting one :razz: .
Similar threads
