I don't know if that's true, but this is a crucial point and the author skips it. Before designing an amp with V or I feedback, he needs to prove, or at least hypothesize as to the intended speaker(s) and the relationship between sound output and V or I.
So basically I'm saying that article is just prose
ElectroVoice provided an adjustment knob for setting the output impedance on their Circlotron Amps. The customer set it to where it sounded best. Seeing as ElectroVoice also made high quality speaker drivers, I would guess they knew what they were doing. But the impedance thing seems to have been a passing fad. Probably didn't make enough difference to be a major selling point.
Assuming the microphone puts out voltage proportional to sound pressure, it would make sense the a Voltage Amp would be called for to convert V back to sound pressure (cone velocity). Your car's dashboard readout indicates Speed rather than torque or HP after all, even though those are what makes it go.
A current amplifier is putting out what amounts to the derivative (time differential) of sound pressure, so it needs it's input signal modified to reflect that. Taking a derivative tends to enhance noise badly, so no surprise that current drive isn't the standard.
Running the standard voltage signal into a current amplifier, without pre-treatment, will give you very boomy bass and a reduced high freq end.
Looks like that's what some are doing anyway.
ElectroVoice on the other hand (and presumeably the "Ultimate" Amp too) was saying that matching the speakers impedance could remove the boomy LF resonance. Totally different from pure current drive.
Assuming the microphone puts out voltage proportional to sound pressure, it would make sense the a Voltage Amp would be called for to convert V back to sound pressure (cone velocity). Your car's dashboard readout indicates Speed rather than torque or HP after all, even though those are what makes it go.
A current amplifier is putting out what amounts to the derivative (time differential) of sound pressure, so it needs it's input signal modified to reflect that. Taking a derivative tends to enhance noise badly, so no surprise that current drive isn't the standard.
Running the standard voltage signal into a current amplifier, without pre-treatment, will give you very boomy bass and a reduced high freq end.
Looks like that's what some are doing anyway.
ElectroVoice on the other hand (and presumeably the "Ultimate" Amp too) was saying that matching the speakers impedance could remove the boomy LF resonance. Totally different from pure current drive.
Last edited:
Yes and if that mike is loaded by a pure resistor, the current is proportional to sound pressure.
No, the readout indicates speed because we have speed limits to observe and obey, not torque limits.
But there ARE cars that allow you to display torque, hp and many other operating parameters. You lost me there. You said the mike gives V, which is proportional to sound pressure p. Now you say I is dp/dt, so the same as dV/dt? Somehow I don't think I=dV/dt.
The speaker drive maintains a cone velocity for a constant voltage input. (standard voltage generator formula, steady flux crossing develops back voltage to counter the drive voltage) Cone velocity translates to sound pressure. So voltage in sets sound pressure.
The voice coil current sets a force for a constant current. F = ma (Newton)
Velocity = integral of acceleration.
You integrate acceleration to get velocity. So a constant current will produce a ramping velocity (and ramping sound pressure). At least until suspension supports start producing enough back force to counter that towards the end of travel.
So yes, I = k dV/dt or dp/dt
A current amplifier needs a differentiator on it's input to derive the correct drive signal for a speaker. (assuming input V is proportional to sound pressure)
Cone support constraints and cabinet back pressure complicate matters further.
The voice coil current sets a force for a constant current. F = ma (Newton)
Velocity = integral of acceleration.
You integrate acceleration to get velocity. So a constant current will produce a ramping velocity (and ramping sound pressure). At least until suspension supports start producing enough back force to counter that towards the end of travel.
So yes, I = k dV/dt or dp/dt
A current amplifier needs a differentiator on it's input to derive the correct drive signal for a speaker. (assuming input V is proportional to sound pressure)
Cone support constraints and cabinet back pressure complicate matters further.
The above discussion is concerning the AC speaker properties. The DC coil resistance is of course directly (in series actually) across the speaker terminals and will greatly effect the actual V and I seen.
Either voltage drive or current drive must produce the same V and I for the same sound. (except possibly for parasitic effects like suspension force or cabinet back pressure)
Taking the constant current case for example, the constant acceleration produces a linear ramping velocity and hence a linear ramping voltage in order to maintain the constant current.
Either voltage drive or current drive must produce the same V and I for the same sound. (except possibly for parasitic effects like suspension force or cabinet back pressure)
Taking the constant current case for example, the constant acceleration produces a linear ramping velocity and hence a linear ramping voltage in order to maintain the constant current.
Last edited:
Smoking:
I'm not worried about theory. I'm worried about real life. If the speaker impedance and efficiency were constant over frequency and amplitude, it wouldn't matter if we use voltage or current. But it's not.
The question that I haven't seen answered is this: For the speaker this guy had in mind, or for any speaker you pick to be driven by your (to be designed) amp, is the sound pressure more closely related to V or I? Again, not at one point, but over the entire frequency and amplitude range of operation.
That's not a question that can be answered theoretically, as far as I can tell. Somebody just needs to take a typical speaker system (whatever "typical" may mean) and run some voltage and current vs pressure curves, sweeping over f and A. Then we'll see if either V or I should be used.
If neither looks good, we'll just have to use a mike to generate the FB signal
I'm not worried about theory. I'm worried about real life. If the speaker impedance and efficiency were constant over frequency and amplitude, it wouldn't matter if we use voltage or current. But it's not.
The question that I haven't seen answered is this: For the speaker this guy had in mind, or for any speaker you pick to be driven by your (to be designed) amp, is the sound pressure more closely related to V or I? Again, not at one point, but over the entire frequency and amplitude range of operation.
That's not a question that can be answered theoretically, as far as I can tell. Somebody just needs to take a typical speaker system (whatever "typical" may mean) and run some voltage and current vs pressure curves, sweeping over f and A. Then we'll see if either V or I should be used.
If neither looks good, we'll just have to use a mike to generate the FB signal
Real speakers are quite complicated when considering radiation patterns and piston size to wavelength ratio, and cabinet effects, room effects etc. One would need either an active Fdbk mike or computer processing to vary the drive parameters in real time. Theory gives you a starting point, then modify, modify....
Because the site worker has spent MUCH time physically scanning dusty copies, and is tired of seeing other sites steal his labor in bulk.
As you say, most PDFs are trivial to crack, but that's apparently enough to reduce leeching.
Apple's docs say you can rotate and do not mention password exception. (I did not hunt for re-size.) On Windows I can rotate and size without trouble. Is Apple doing it some different way?? (I am not asking you; only Apple knows, and may not know they know.)
Ha. I had not noticed that I can not PRINT these files. Mildly limiting indeed.
FreeMyPDF.com - Removes passwords from viewable PDFs
"Viewable PDFs with restrictions are not really protected, no more than a door with a broken lock, as opposed to password-protected PDFs."
Indeed, that site removed the no-print limit on that file. Also copying:
"The ULTIMATE anni/iiki by EMORY COOK and GUS JOSE
"If the sentiments of Walter Mitty are sometimes yours, here's a golden opportunity to live dangerously for awhile - build this super amplifier, whose design is explained in a refreshing manner by the perpetrators."
("anni/iiki" is "amplifier in a script font and OCR failure is fine.)
Sometimes the missing part that trips folks up, to the point of semantic arguments, is that speaker motors don't operate in "real time". A magnetic field is established at time t and the moving system moves to attempt to zero this induced field against the magnet's fixed field. It's always behind the force at time t, moving with smoking-amp's ramping velocity.
Magnetic microphones and speakers are said to be mass controlled when they're velocity sensitive and mass integrated.
All good fortune,
Chris
A ribbon microphone works on the velocity of air.
Air velocity from either in front of, or behind the back of the microphone produces an output signal.
Air velocity that passes from one side of the microphone does not produce an output signal.
And a pressure wave that comes from the side of the ribbon microphone does not produce an output signal.
The ribbon microphone polar pickup pattern is a figure 8.
In much the same way, a ribbon loudspeaker is a velocity device, with a figure 8 polar pattern.
Standing Waves in a room:
A closed box woofer speaker produces a pressure wave, the polar pattern is in all directions.
In a room, it activates 3 Axis: Width, Depth, and Height dimensions of the room.
On the other hand . . .
An open baffle woofer speaker produces a velocity wave, and it only activates 1 of the 3 Axis that are mentioned above.
If you have ever heard the bass of a good open baffle, versus the bass of a good closed box, you will easily be able to hear the difference.
There are only 1/3 of the bass standing wave frequencies for the open baffle, versus the close box standing wave frequencies.
Air velocity from either in front of, or behind the back of the microphone produces an output signal.
Air velocity that passes from one side of the microphone does not produce an output signal.
And a pressure wave that comes from the side of the ribbon microphone does not produce an output signal.
The ribbon microphone polar pickup pattern is a figure 8.
In much the same way, a ribbon loudspeaker is a velocity device, with a figure 8 polar pattern.
Standing Waves in a room:
A closed box woofer speaker produces a pressure wave, the polar pattern is in all directions.
In a room, it activates 3 Axis: Width, Depth, and Height dimensions of the room.
On the other hand . . .
An open baffle woofer speaker produces a velocity wave, and it only activates 1 of the 3 Axis that are mentioned above.
If you have ever heard the bass of a good open baffle, versus the bass of a good closed box, you will easily be able to hear the difference.
There are only 1/3 of the bass standing wave frequencies for the open baffle, versus the close box standing wave frequencies.
Last edited:
Ribbons are flattened and exposed coils (that are often also their own diaphragm), in a fixed magnetic field, so have the same motor characteristics as a "dynamic" microphone or speaker. Ordinary dynamic speakers will have the same patterns if similarly exposed. These generators / drivers would have a 20dB per decade rising magnitude response, like a magnetic phono cartridge does, except that we use them above their fundamental mass resonance, which adds a 20dB per decade falling responce. Another definition of "mass controlled".
Electrostatic microphones have constant amplitude responce, so must operate below their fundamental resonance.
Some dynamic microphones, like the Shure SM-57, etc. have resonances in middle of their passband. Hot mess.
All good fortune,
Chris
Electrostatic microphones have constant amplitude responce, so must operate below their fundamental resonance.
Some dynamic microphones, like the Shure SM-57, etc. have resonances in middle of their passband. Hot mess.
All good fortune,
Chris
Last edited:
I want to thank you guys very much for producing this feed and getting me this article. I bought a pair of these amplifiers about 10 years ago, and I have been looking for the original article since. Thanks for everything. I actually was about ready to part them out today, until I found this article. Definitely going to keep them intact. I did have a listen to them when I first got them, and they sounded pretty good. Now I am going to go over each of them with a microscope to see if they need anything. Thanks again,
A.C.
A.C.
Philosophically I note that nearly every ultimate eventually becomes the penultimate.
I like the idea of current drive but for multiway speakers it seems the crossover always mucks that up to some degree. I am also not a big fan of including the OPT in the loop (I know it is the majority report) as it seems to be trying to beat a rabid dog into submission. You can easily get some wild push back.
I like the idea of current drive but for multiway speakers it seems the crossover always mucks that up to some degree. I am also not a big fan of including the OPT in the loop (I know it is the majority report) as it seems to be trying to beat a rabid dog into submission. You can easily get some wild push back.