I realize that a current-source amp doesn't care about the impedance in that the magnitude of the current delivered is indepedent of the impedance; however, what about the phase? The current across an inductor lags behind the voltage, so from a conventional voltage-source amp the motion of the speaker would lag behind the voltage waveform going across its voice coil (I think).
Is it the same way with a current-source amp?
Is it the same way with a current-source amp?
454Casull said:
If you drive a loudspeaker with an ideal current source, the voice coil inductance and resistance does not affect the acoustic output at all. I think that answers your question?
With the normal near-ideal voltage source the output is greatly affected by both. The voice coil inductance acts as a lowpass filter, typically of order 0.7 or so*, and the resistance keeps the response from peaking at fs (the resonant frequency of the driver).
The resistance effect is highly desirable, but the inductance might not be.
*Yes, filters are not only of integer orders. Since the voice coil inductance has an impedance that typically is proportional to f^(0.7) one could say that the filter order becomes 0.7.
Re: Re: Re: Impedance and current-source amplification
Yes. No. The inductance is still the same, but if the inductance varies, the output voltage of the current source will vary to compensate for that. Imagine that an ideal current source has an infinite output impedance. You can connect any impedance in series with that, and the output impedance will still be infinite, ie no change.
This has another positive effect; if the inductance is non-linear and causes distortion in the constant-voltage drive condition, it causes NO distortion in the constant-current drive condition. Funnily, the current source will produce a distorted voltage and this distortion compensates for the distortion caused by the non-linear inductance.
454Casull said:
Yes. No. The inductance is still the same, but if the inductance varies, the output voltage of the current source will vary to compensate for that. Imagine that an ideal current source has an infinite output impedance. You can connect any impedance in series with that, and the output impedance will still be infinite, ie no change.
This has another positive effect; if the inductance is non-linear and causes distortion in the constant-voltage drive condition, it causes NO distortion in the constant-current drive condition. Funnily, the current source will produce a distorted voltage and this distortion compensates for the distortion caused by the non-linear inductance.
Okay, that might be true for an ideal current source amp.
Not that I am any expert on this, (far from it), but I am given to understand that in real life current source amps the output will vary quite a bit with the impedance variation. This is in direct contrast with most solid state amps with near ideal voltage source behavior.
Is this true?
Not that I am any expert on this, (far from it), but I am given to understand that in real life current source amps the output will vary quite a bit with the impedance variation. This is in direct contrast with most solid state amps with near ideal voltage source behavior.
Is this true?
kelticwizard said:
Since Vout = Iout * Rload, with current amps, Iout is fixed for a specific level, Rload changes therefore Vout varies all over the place. But the idea is that Vout doesn't matter with current drive, it's the current that matters. And that is a fixed ratio to input signal.
So, yes, it's true, but who cares?
Jan Didden
kelticwizard said:
No.
You can make a very-near ideal current source. It will of course have the same limitations regarding maximum output voltage and current as the typical near-ideal voltage source (=most normal amplifiers).
Here is a proof-of-concept design:
An externally hosted image should be here but it was not working when we last tested it.
454Casull said:
I think you are right, but I am not sure. I wouldn't mind some verification on that myself.
I do remember reading in Audio magazine an interview with a prominent designer, (I forget his name). He said that he thought in a few years, the cone loudspeaker would disappear from the high end of the audio market because in one octave, the signal comes at you as if it is ten feet in front of you, but in another octave the signal comes at you as if it was much farther back. I think that might be what he was referring to, but once again I am not entirely certain.
Jan and Svante:
I think I botched up the wording of the post you both replied to.
When I said:
I was thinking of the SPL from a traditional moving coil speaker or speaker system , not the output the amp delivers to the speaker. If a speaker has an irregular impedance curve but a smooth SPL output when hooked up to a solid state voltage driven amp, I am given to understand that when the same speaker is hooked up to a current driven amp the SPL will vary with the impedance plot.
Am I correct on this?
I think I botched up the wording of the post you both replied to.
When I said:
I was thinking of the SPL from a traditional moving coil speaker or speaker system , not the output the amp delivers to the speaker. If a speaker has an irregular impedance curve but a smooth SPL output when hooked up to a solid state voltage driven amp, I am given to understand that when the same speaker is hooked up to a current driven amp the SPL will vary with the impedance plot.
Am I correct on this?
kelticwizard said:
Ah, I see! Yes, that is the big problem with current drive; it will peak at the system resonance(s). Effectively, what happens is that Qts of the driver becomes Qms and Qes is made completely unimportant.
The output towards higher frequencies will increase too, due to the voice coil inductance.
Okay, just a couple of more questions.
Below is the response plot at 2.83 volts from a constant voltage amp, of a Peerless PLW-009-8, formerly of the Vifa line. No Faraday ring on this model. As you can see, the impedance starts to rise around 600 Hz and goes up fairly sharply.
Although we do not have a phase diagram, just an impedance graph, I think we can safely assume the rise in impedance is due to the speaker becoming more inductive at 7 KHz than at 600 Hz. The impedance seems to be around 8 ohms at 600 Hz, over 20 ohms at 7 KHz.
A) If this speaker is driven from a constant voltage amp, as it is in the illustration, will the sound at, say 7 KHz be delayed in respect to the sound at 600 Hz?
B) If this same speaker is driven from a constant current amp, will the sound be delayed at 7 KHz compared to 600 Hz?.
Below is the response plot at 2.83 volts from a constant voltage amp, of a Peerless PLW-009-8, formerly of the Vifa line. No Faraday ring on this model. As you can see, the impedance starts to rise around 600 Hz and goes up fairly sharply.
Although we do not have a phase diagram, just an impedance graph, I think we can safely assume the rise in impedance is due to the speaker becoming more inductive at 7 KHz than at 600 Hz. The impedance seems to be around 8 ohms at 600 Hz, over 20 ohms at 7 KHz.
A) If this speaker is driven from a constant voltage amp, as it is in the illustration, will the sound at, say 7 KHz be delayed in respect to the sound at 600 Hz?
B) If this same speaker is driven from a constant current amp, will the sound be delayed at 7 KHz compared to 600 Hz?.
Attachments
kelticwizard said:
First I'd like to state that the word delay is somewhat misleading in this case. There will be phase shifts, yes, but the transition from phase shift to delay is not easy. Let's say we have a sinusoid at 1 kHz and a second sinusoid that is shifted 90 degrees towards a later time. Is the delay 0,25 ms? Is it 1,25 ms? or is it possibly -0,75 ms? The point here is that if we talk about individual frequencies, there is no onset of the signal and so we cannot determine the delay at a single frequency. We can convert the phase shift to group delay or even phase delay, but neither is easily translated to the type of delay that occurs if we move away from a loudspeaker, for example. The wrod delay is simply not well defined for single frequencies.
So let's talk about phase shifts.
Second, since we are talking here about the phase shifts that are potentially introcuced by using current/voltage drive it is easier to understand if we think of a more idealized loudspeaker, skip the jaggedness of the curve towards higher frequencies; these come from cone breakups.
If we do so, and drive them with either current or voltage drive we get the curves below:
An externally hosted image should be here but it was not working when we last tested it.
Red is constant voltage, blue is constant current. We can see that there are changes in phase (middle graph) both around the resonance near 50 Hz and towards higher frequencies. If we convert the phase differences to group delay (which I am personally not very fond of) we can see (upper graph) that the delay is rediculously small compared to that near the resonance.
If we take the real case that you linked to, the difference between current and voltage drive will be the same, only that the phase shift that the cone breakups are added.
So, to attempt to answer your questions: There will be phase shifts in both cases, and the voltage drive case will have a more negative phase shift than the current drive, since the inductance is effectively out out of the picture with current drive. If you would play 7 kHz on two identical systems, one with current drive and one with voltage drive, the voltage driven system would lag the current driven.
...but regardless of the above the changes in the frequency response (the amplitude) is by far the most audible. There is no way that you could compare the two systems by listening to them and state that either one is better because of delay issues. The FR effects is far more perceptually important.
Hello,
Participants to the recent European Triode Festival could attend to a very intersting conference given by Pete Millet about driving a loudspeaker with voltage, given output impedance or current (= very high Zout).
Pete Millet did to us some demonstrations using a special amplifier he built specially for the conference.
With the use of a potentiometer the output impedance of this amplifier can be set ,for a constant volume output, to any Zout between 0 and 500ohms.
Demonstrations were done using 3 different Zout: 0ohm, 16ohms and 500ohms on 3 different loudspeakers (in mono)
1) a very small 2 ways enclosure
2) a bicone full-range loudspeaker in a medium sized box
3) a fostex 206 without the inner cone loaded by a horn having 1meter diameter (4inches diameter) unfiltered.
On 1) it was difficult to notice a change, may be the bass was too mudy using Zout = 500ohms.
On 2) there was some noticeable change and a majority of participants preferred a 16ohms value for Zout.
On 3) there was also some notieable changes, rising the output impedance to 500ohms , the lowmid appeared to be more coherent in position than the medium and the treble (with Zout = 0 low-mid seems to comes from the depth in the horn) and the image was far better. Also the general balance of the response was far better.
So large impedance seems to be very benefitial to some loudspeakers (give a look to Nelson Pass for additional evidences). We have to keep in mind that loudspeakers initially are current control device. When their acoustic load is resistive it is desireable to control them by current. It is also very desireable to control them in current when their response is the inverse of their impedance curve (like with the TAD TD2001 on good horns which reponse is perfectly linearized using current drive)...
The problem arrise when trying to use them with reactive load (in order to enlarge their bandwith, like with close enclosures or bass-reflex...) as optimisation of such load is based on a constant voltage hypothesis.
About phase shift when control is in voltage or in current, there is a always orgotten thing: the acoustical pressure at the loudspeaker output is in quadrature with the signal input (so probably the notion of "phase distortion" applies).
reference: (Citation)
"LOCATION OF SOUND SOURCES"
BENJAMIN B. DRISKO
Science Industry Foundation, Rockport, Maine
Interest in a more exact location of a sound source led to the following experiment. With a sound source (JBL tweeter 175DLH) and a microphone (640AA) in an anechoic chamber, a Lissajou pattern was formed on the scope by the voltage feeding the source vs the output voltage from the microphone. A list of frequencies for which the phase was either zero or 180° was recorded.
A few seconds work with this raw data indicated that no results would be forthcoming without accurate knowledge of the velocity of sound. It was then decided that the addition of a second similar list taken from a different but accurately known distance would, of necessity, lead to the answer.
Thus, a plot was constructed on linear rectangular coordinate paper of the frequencies versus the half-wave numbers. Except for second and higher order effects, these points were found to lie on a straight line.
Extrapolating this line back to zero frequency, it was seen to intercept the wave number axis, not at an integral number but at half of a half-wavelength or a quarter wavelength.
At first thought this seemed utterly unthinkable. The reasons are not completely clear yet, although the fact seems established beyond all doubt.
If one plots frequency as ordinate and wavelength as abscissa, then the slope of the "best fit" straight line is cycles per wavelength.
In other words, this quotient is the frequency whose distance from source to mike is exactly one wavelength, and the reciprocal of this is precisely the transit time from source to mike. Since the accurately known difference between two unknown distances divided by the known time difference is the velocity of sound, the above can be used to locate the sound source: the velocity of sound times the transit time is the distance from mike to source. This result never comes anywhere near the location of the loudspeaker.
The first analysis consisted of making a plot as above and fitting a straight line by the "eyeball method." There was no question but that the line simply would not go through either zero or one-half wavelength.
If all wave numbers are divided by corresponding frequencies, there results a table of transit times which will be uniform if the wave numbers are correct and will have increasing errors for low numbers if the wave numbers are wrong. High wave numbers will be asymptotic to the correct time in both cases. By this means, performing several trials and errors, the true wave numbers were ascertained. This method involved no eyeball gymnastics and resuited in the answer that at zero frequency the wave number is 0,25.
At this point the original plot was redrawn at a much larger scale and with the bottom end of the line locked at the point [0 ; 0,25]. It was found in this manner that the sound velocity was 1129 ft/sec and that the sound was actually originating at a point 1,1 inch behind the housing or about 1,5 inch behind the diaphragm. The overall accuracy appears to be approximately 10 microseconds or 0,1 inch.
Search through about a dozen acoustical texts failed to find any reference either to the method or the findings.
"SOME PHASE CONSIDERATIONS IN SOUND LOCATING PROCEDURE"
F.K. HARVEY
Bell Telephone Laboratories, Inc., Murray Hill, New Jersey
In a recent Letter to the Editor on "Location of Sound Sources" by Benjamin B. Drisko (J.A.E.S. 12, 166) that author discusses some experimental data involving frequency vs wavelengths-to-source at a point in free space in front of a loudspeaker. His measurements were made by comparing the phase of the electrical input at the loudspeaker with the electrical output of the pickup microphone. This information was converted to wavelengths-to-source at the frequencies observed. A straight line plot of the data was described as consistently having an intercept of 0,25 wavelength at zero frequency. From this the apparent sound source was calculated to be behind the loudspeaker diaphragm.
It should be pointed out that, although the measuring microphone is generally free of phase distortion, the ordinary direct radiator loudspeaker is not. The latter is generally operated above resonance (mass controlled), and thus provides a nominal 90° phase shift (0,25 wavelength) at most frequencies throughout the useful range. In such a loudspeaker, the mechanical driving force is in phase with the electrical current. However, the diaphragm velocity is 90° out of phase with this force and its magnitude drops 6 db/octave. It is purposely designed in this way so that the product of the velocity squared and the air radiation resistance (rising 12 dB/octave) tends to provide a constant acoustical power output up to the point where the dimensions of the diaphragm become comparable with the wavelength.
It is generally overlooked that one of our most popular loudspeaker designs possesses an inherent 90° phase distortion. Consequently, preservation of the complex waveform is impossible, even in free space, except with the aid of preliminary amplitude-independent phase correction. In view of this situation, it might be better to plot isophase contours in space coordinates in front of the loudspeaker. At any particular frequency the apparent location of the source can be established from the center of curvature of the phase contours in the region of interest.
REFERENCES:
M. Rettinger, Practical Electroacoustics (Chemical Publishing Company,
New York, 1955), pp. 61-66.
W. E. Kock and F. K. Harvey, "A Photographic Method of Displaying
Sound Wave and Microwave Space Patterns," Bell System
Tech. J. 30, 564 (July, 1951).
Best regards from Paris, France
Jean-Michel Le Cléac'h
Participants to the recent European Triode Festival could attend to a very intersting conference given by Pete Millet about driving a loudspeaker with voltage, given output impedance or current (= very high Zout).
Pete Millet did to us some demonstrations using a special amplifier he built specially for the conference.
With the use of a potentiometer the output impedance of this amplifier can be set ,for a constant volume output, to any Zout between 0 and 500ohms.
Demonstrations were done using 3 different Zout: 0ohm, 16ohms and 500ohms on 3 different loudspeakers (in mono)
1) a very small 2 ways enclosure
2) a bicone full-range loudspeaker in a medium sized box
3) a fostex 206 without the inner cone loaded by a horn having 1meter diameter (4inches diameter) unfiltered.
On 1) it was difficult to notice a change, may be the bass was too mudy using Zout = 500ohms.
On 2) there was some noticeable change and a majority of participants preferred a 16ohms value for Zout.
On 3) there was also some notieable changes, rising the output impedance to 500ohms , the lowmid appeared to be more coherent in position than the medium and the treble (with Zout = 0 low-mid seems to comes from the depth in the horn) and the image was far better. Also the general balance of the response was far better.
So large impedance seems to be very benefitial to some loudspeakers (give a look to Nelson Pass for additional evidences). We have to keep in mind that loudspeakers initially are current control device. When their acoustic load is resistive it is desireable to control them by current. It is also very desireable to control them in current when their response is the inverse of their impedance curve (like with the TAD TD2001 on good horns which reponse is perfectly linearized using current drive)...
The problem arrise when trying to use them with reactive load (in order to enlarge their bandwith, like with close enclosures or bass-reflex...) as optimisation of such load is based on a constant voltage hypothesis.
About phase shift when control is in voltage or in current, there is a always orgotten thing: the acoustical pressure at the loudspeaker output is in quadrature with the signal input (so probably the notion of "phase distortion" applies).
reference: (Citation)
"LOCATION OF SOUND SOURCES"
BENJAMIN B. DRISKO
Science Industry Foundation, Rockport, Maine
Interest in a more exact location of a sound source led to the following experiment. With a sound source (JBL tweeter 175DLH) and a microphone (640AA) in an anechoic chamber, a Lissajou pattern was formed on the scope by the voltage feeding the source vs the output voltage from the microphone. A list of frequencies for which the phase was either zero or 180° was recorded.
A few seconds work with this raw data indicated that no results would be forthcoming without accurate knowledge of the velocity of sound. It was then decided that the addition of a second similar list taken from a different but accurately known distance would, of necessity, lead to the answer.
Thus, a plot was constructed on linear rectangular coordinate paper of the frequencies versus the half-wave numbers. Except for second and higher order effects, these points were found to lie on a straight line.
Extrapolating this line back to zero frequency, it was seen to intercept the wave number axis, not at an integral number but at half of a half-wavelength or a quarter wavelength.
At first thought this seemed utterly unthinkable. The reasons are not completely clear yet, although the fact seems established beyond all doubt.
If one plots frequency as ordinate and wavelength as abscissa, then the slope of the "best fit" straight line is cycles per wavelength.
In other words, this quotient is the frequency whose distance from source to mike is exactly one wavelength, and the reciprocal of this is precisely the transit time from source to mike. Since the accurately known difference between two unknown distances divided by the known time difference is the velocity of sound, the above can be used to locate the sound source: the velocity of sound times the transit time is the distance from mike to source. This result never comes anywhere near the location of the loudspeaker.
The first analysis consisted of making a plot as above and fitting a straight line by the "eyeball method." There was no question but that the line simply would not go through either zero or one-half wavelength.
If all wave numbers are divided by corresponding frequencies, there results a table of transit times which will be uniform if the wave numbers are correct and will have increasing errors for low numbers if the wave numbers are wrong. High wave numbers will be asymptotic to the correct time in both cases. By this means, performing several trials and errors, the true wave numbers were ascertained. This method involved no eyeball gymnastics and resuited in the answer that at zero frequency the wave number is 0,25.
At this point the original plot was redrawn at a much larger scale and with the bottom end of the line locked at the point [0 ; 0,25]. It was found in this manner that the sound velocity was 1129 ft/sec and that the sound was actually originating at a point 1,1 inch behind the housing or about 1,5 inch behind the diaphragm. The overall accuracy appears to be approximately 10 microseconds or 0,1 inch.
Search through about a dozen acoustical texts failed to find any reference either to the method or the findings.
"SOME PHASE CONSIDERATIONS IN SOUND LOCATING PROCEDURE"
F.K. HARVEY
Bell Telephone Laboratories, Inc., Murray Hill, New Jersey
In a recent Letter to the Editor on "Location of Sound Sources" by Benjamin B. Drisko (J.A.E.S. 12, 166) that author discusses some experimental data involving frequency vs wavelengths-to-source at a point in free space in front of a loudspeaker. His measurements were made by comparing the phase of the electrical input at the loudspeaker with the electrical output of the pickup microphone. This information was converted to wavelengths-to-source at the frequencies observed. A straight line plot of the data was described as consistently having an intercept of 0,25 wavelength at zero frequency. From this the apparent sound source was calculated to be behind the loudspeaker diaphragm.
It should be pointed out that, although the measuring microphone is generally free of phase distortion, the ordinary direct radiator loudspeaker is not. The latter is generally operated above resonance (mass controlled), and thus provides a nominal 90° phase shift (0,25 wavelength) at most frequencies throughout the useful range. In such a loudspeaker, the mechanical driving force is in phase with the electrical current. However, the diaphragm velocity is 90° out of phase with this force and its magnitude drops 6 db/octave. It is purposely designed in this way so that the product of the velocity squared and the air radiation resistance (rising 12 dB/octave) tends to provide a constant acoustical power output up to the point where the dimensions of the diaphragm become comparable with the wavelength.
It is generally overlooked that one of our most popular loudspeaker designs possesses an inherent 90° phase distortion. Consequently, preservation of the complex waveform is impossible, even in free space, except with the aid of preliminary amplitude-independent phase correction. In view of this situation, it might be better to plot isophase contours in space coordinates in front of the loudspeaker. At any particular frequency the apparent location of the source can be established from the center of curvature of the phase contours in the region of interest.
REFERENCES:
M. Rettinger, Practical Electroacoustics (Chemical Publishing Company,
New York, 1955), pp. 61-66.
W. E. Kock and F. K. Harvey, "A Photographic Method of Displaying
Sound Wave and Microwave Space Patterns," Bell System
Tech. J. 30, 564 (July, 1951).
Best regards from Paris, France
Jean-Michel Le Cléac'h
Jmmlc said:
Ok... I bet that 99% of the perceived differences can be explained with the (amplitude) frequency response changes that resulted from the impedance change. But maybe you don't mean anything else?
Jmmlc said:
Hmm, by "in quadrature" do you mean 90 degrees out of phase?
Then I definitely have a different opinion, sound pressure is in phase with the driving voltage, given that the speaker has a flat response and that the delay between the speaker and mic is neglected.
I don't understand what the references has to do with the topic. As far as I understand (but I might be wrong) they consider phase differences between channels which is something completely different from phase differences between frequencies.
Hello,
About the reference:
There is much more that it seems at the first glance and NO the measurement is not about interchannel as the measurement is done in mono on a single loudspeaker.
The procedure used was developped to know precisely where the source of the sound is positionned (at the coil somewhere on the diaphragm...etc). The result is that the source is a quarter wavelength behind the coil of the loudspeaker.
Location of Sound Sources
JAESVolume 12 Number 2 p. 166; April 1964
Author: Drisko, Benjamin B.
E-lib Location: (CD aes2) /jrnl5367/1964/6511.pdf
The reply from Harvey (Bell lab) has never been objected.
Some Phase Considerations in Sound Locating Procedure
JAES Volume 12 Number 4 p. 348; October 1964
Author: Harvey, F. K.
E-lib Location: (CD aes2) /jrnl5367/1964/6535.pdf
About current control of a loudspeaker, I don't have the refernce handy but there was in JAES a paper indicating that current control of a loudspeaker can result in a dramatic decrease in distortion in the medium...
Best regards from Paris.
Jean-Michel Le Cléac'h
About the reference:
There is much more that it seems at the first glance and NO the measurement is not about interchannel as the measurement is done in mono on a single loudspeaker.
The procedure used was developped to know precisely where the source of the sound is positionned (at the coil somewhere on the diaphragm...etc). The result is that the source is a quarter wavelength behind the coil of the loudspeaker.
Location of Sound Sources
JAESVolume 12 Number 2 p. 166; April 1964
Author: Drisko, Benjamin B.
E-lib Location: (CD aes2) /jrnl5367/1964/6511.pdf
The reply from Harvey (Bell lab) has never been objected.
Some Phase Considerations in Sound Locating Procedure
JAES Volume 12 Number 4 p. 348; October 1964
Author: Harvey, F. K.
E-lib Location: (CD aes2) /jrnl5367/1964/6535.pdf
About current control of a loudspeaker, I don't have the refernce handy but there was in JAES a paper indicating that current control of a loudspeaker can result in a dramatic decrease in distortion in the medium...
Best regards from Paris.
Jean-Michel Le Cléac'h
Hello,
Here is the 2 references of the papers about distortion reduction in a loudspeaker driven by current.
Transconductance Power Amplifier Systems for Current-Driven Loudspeakers
Authors: Mills, P. G. L.; Hawksford, M. O. J.
JAES Volume 37 Number 10 pp. 809-822; October 1989
E-lib Location: (CD aes5) /jrnl8997/1989/8302.pdf
"Moving-coil loudspeakers generally provide a substantial improvement in linearity when current driven, together with the elimination of voice-coil heating effects. Consequently there is a need to investigate low-distortion power amplifier topologies suitable for this purpose. After considering established current feedback approaches, a novel method using a common-base isolation stage is outlined and extended to show a prototype amplifier circuit in detail. In addition, the elements of a two-way active current-driven system are described, with low-frequency velocity feedback control derived from a sensing coil. The coupling error between this coil and the main driving coil is nulled by electronic compensation."
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Distortion Reduction in Moving-Coil Loudspeaker Systems Using Current-Drive Technology
Authors: Mills, P. G. L.; Hawksford, M. O. J.
JAES Volume 37 Number 3 pp. 129-148; March 1989
E-lib Location: (CD aes5) /jrnl8997/1989/8269.pdf
"The performance advantages of current-driving moving-coil loudspeakers is considered, thus avoiding thermal errors caused by voice-coil heating, nonlinear electromagnetic damping due to (Bl)2 variations, and high-frequency distortion from coil inductive effects, together with reduced interconnect errors. In exploring methods for maintaining system damping, motional feedback is seen as optimal for low-frequency applications, while other methods are considered. The case for current drive is backed by nonlinear computer simulations, measurements, and theoretical discussion. In addition, novel power amplifier topologies for current dirve are discussed, along with methods of drive-unit thermal protection. "
Best regards from Paris,
Jean-Michel Le Cléac'h
Here is the 2 references of the papers about distortion reduction in a loudspeaker driven by current.
Transconductance Power Amplifier Systems for Current-Driven Loudspeakers
Authors: Mills, P. G. L.; Hawksford, M. O. J.
JAES Volume 37 Number 10 pp. 809-822; October 1989
E-lib Location: (CD aes5) /jrnl8997/1989/8302.pdf
"Moving-coil loudspeakers generally provide a substantial improvement in linearity when current driven, together with the elimination of voice-coil heating effects. Consequently there is a need to investigate low-distortion power amplifier topologies suitable for this purpose. After considering established current feedback approaches, a novel method using a common-base isolation stage is outlined and extended to show a prototype amplifier circuit in detail. In addition, the elements of a two-way active current-driven system are described, with low-frequency velocity feedback control derived from a sensing coil. The coupling error between this coil and the main driving coil is nulled by electronic compensation."
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Distortion Reduction in Moving-Coil Loudspeaker Systems Using Current-Drive Technology
Authors: Mills, P. G. L.; Hawksford, M. O. J.
JAES Volume 37 Number 3 pp. 129-148; March 1989
E-lib Location: (CD aes5) /jrnl8997/1989/8269.pdf
"The performance advantages of current-driving moving-coil loudspeakers is considered, thus avoiding thermal errors caused by voice-coil heating, nonlinear electromagnetic damping due to (Bl)2 variations, and high-frequency distortion from coil inductive effects, together with reduced interconnect errors. In exploring methods for maintaining system damping, motional feedback is seen as optimal for low-frequency applications, while other methods are considered. The case for current drive is backed by nonlinear computer simulations, measurements, and theoretical discussion. In addition, novel power amplifier topologies for current dirve are discussed, along with methods of drive-unit thermal protection. "
Best regards from Paris,
Jean-Michel Le Cléac'h
Jmmlc said:
If this is the case, then if you have a 5 inch speaker which covers the range of 100 Hz up to 3,000 Hz, (not an unusual situation in this era of very common subs with their own crossover/amp combos built in), then at 100 Hz the sound source is 33.75 inches behind the voice coil, but at 3,000 Hz the sound source is 1.125 inches behind the voice coil. The high frequency sound source is located 32.625 inches closer than the source of the low frequency.
Would you agree?
Hello,
When I study my own loudspeakers phase measurements versus frequency in order to calculate a phase delay (Phi/omega) (not the group delay) I very rarely find value of N.Pi for the estimated phase at frequency 0Hz. Most often the difference to N.Pi is something around Pi/2.
The only explanation available I found is the JAES papers I mentionned. (While Drisko's measurements were done on a JBL tweeter, Harvey's reply affirms the phenomenon is universal!)
I have no opinion on that myself I would want other opinions but a demonstration should be better.
How this should lead to a difference between voltage drive and current drive is an interesting question too.
Best regards from Paris,
Jean-Michel Le Cléac'h
When I study my own loudspeakers phase measurements versus frequency in order to calculate a phase delay (Phi/omega) (not the group delay) I very rarely find value of N.Pi for the estimated phase at frequency 0Hz. Most often the difference to N.Pi is something around Pi/2.
The only explanation available I found is the JAES papers I mentionned. (While Drisko's measurements were done on a JBL tweeter, Harvey's reply affirms the phenomenon is universal!)
I have no opinion on that myself I would want other opinions but a demonstration should be better.
How this should lead to a difference between voltage drive and current drive is an interesting question too.
Best regards from Paris,
Jean-Michel Le Cléac'h
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.