: A new Method for Controlling Low-Frequency Loudspeaker Behaviour
$ 20 for non AES members.
Worth buying? Does it present a practical evaluation circuit?
I can see the merit for high Qes drivers forced electronicalyl to be low Qes,but is there some disadvantage apart from complexity?
Ive seen the experimental schematic as available in HIGH PERFORMANCE LOUDSPEAKERS edited by MARTIN COLLOMS.
Seems interesting to adjust amplifier output impedence to negate Loudspeaker Re,but then also adjust CMS etc(directly or indirectly?)
Im interested in both servo and this issue wrt transient response and resonance of bass units.
http://www.aes.org/publications/preprints/preprints_search.cfm
Mike.e
http://www.1388.com/articles/tech_Ace-Bass/
$ 20 for non AES members.
Worth buying? Does it present a practical evaluation circuit?
I can see the merit for high Qes drivers forced electronicalyl to be low Qes,but is there some disadvantage apart from complexity?
Ive seen the experimental schematic as available in HIGH PERFORMANCE LOUDSPEAKERS edited by MARTIN COLLOMS.
Seems interesting to adjust amplifier output impedence to negate Loudspeaker Re,but then also adjust CMS etc(directly or indirectly?)
Im interested in both servo and this issue wrt transient response and resonance of bass units.
http://www.aes.org/publications/preprints/preprints_search.cfm
Mike.e
http://www.1388.com/articles/tech_Ace-Bass/
The AES papers are horrendously expensive. For a little more than $ 20, you can get the full AES anthologies from AudioXpress, that's a lot of good reading.
The Stahl idea is described in a patent you can find at Free Patents On Line. The AES paper is not very different. Have a look to other patents using similar ideas, too (Barbetta, Hahne, Nagi, Saville, Walker, Massao, Yamaha...)
The Stahl patent is in fact partly based on Voigt ideas which are anterior to Black's (negative feedback) and Rice and Kellog's (electrondynamic loudspeaker) patents.
I agree with all Mile.E. wrote
Stahl's basic idea is as genious as simple.
An amplifer and the loudspeaker connected to it must be considered a whole, not as two separate entities. This is evident when one considers that loudspeakers parameters are electro-mechanical parameters.
In Stahl and similar embodiments, the electronics circuit can be considered as having two imbricated parts. A first one is an amplifer with a standard negative feedback loop and a with second feedback loop, positive, which cancels the Re of the loudspeakers by the effect of negative impedance. Then we can consider we have a direct access to the Les, Res and Ces of the loudspeaker. Voltage directly applied to this impedance defines the cone velocity.
A second loop around the amplifier and using op-amps at its input defines its own complex impedance which comes in parallel with the loudspeaker one
(remember : a perfect voltage amplifier has zero output impedance, it can be considered as a short circuit generating voltage) . Then the electro-mechanical parameters of the whole system can be modified at will, in certain limits of course.
Some details to look after :
- the resistor sensing the current to generate the negative impedance should have the same temperature sensitivity as the voice coil. It seems this has been very well engineered by Yamaha.
- just as in motion feedback systems, an important problem is the inductance of the voice coil, Le, and its cancellation : its value is not constant with frequency and with the position inside the gap.
Negative resistance of amplifiers is obtained by positive feedback. It is very easy to obtain it, without using an additionnal op-amp, by bootstraping with adequate resistors the input of the amplifier with the voltage accross the sensing resistor (overcurrent precautions are mandatory).
I have a strange point of view about negative resistance amplifiers. From the amplifier input, there is positive feedback, yes. But viewed from the impedance of the loudspeaker, it can be considered a negative feedback : the current through it decreases as its impedance increases. Then I tend to consider Stahl system as a kind of motion feedback system, disagreeing with his comment who said it is not. This is not an important question.
By controling the impedance of the amplifier, you can much extend the possibilities of a given loudspeaker or adjust it more adequately to your needs. Because it deals with the essential parameters of loudspeakers, I think it is a more interesting area to investigate than the endless discussion about the sound of amplifiers and the supposed horrible effects of feedback.
~~~~~~ Forr
§§§
The Stahl idea is described in a patent you can find at Free Patents On Line. The AES paper is not very different. Have a look to other patents using similar ideas, too (Barbetta, Hahne, Nagi, Saville, Walker, Massao, Yamaha...)
The Stahl patent is in fact partly based on Voigt ideas which are anterior to Black's (negative feedback) and Rice and Kellog's (electrondynamic loudspeaker) patents.
I agree with all Mile.E. wrote
Stahl's basic idea is as genious as simple.
An amplifer and the loudspeaker connected to it must be considered a whole, not as two separate entities. This is evident when one considers that loudspeakers parameters are electro-mechanical parameters.
In Stahl and similar embodiments, the electronics circuit can be considered as having two imbricated parts. A first one is an amplifer with a standard negative feedback loop and a with second feedback loop, positive, which cancels the Re of the loudspeakers by the effect of negative impedance. Then we can consider we have a direct access to the Les, Res and Ces of the loudspeaker. Voltage directly applied to this impedance defines the cone velocity.
A second loop around the amplifier and using op-amps at its input defines its own complex impedance which comes in parallel with the loudspeaker one
(remember : a perfect voltage amplifier has zero output impedance, it can be considered as a short circuit generating voltage) . Then the electro-mechanical parameters of the whole system can be modified at will, in certain limits of course.
Some details to look after :
- the resistor sensing the current to generate the negative impedance should have the same temperature sensitivity as the voice coil. It seems this has been very well engineered by Yamaha.
- just as in motion feedback systems, an important problem is the inductance of the voice coil, Le, and its cancellation : its value is not constant with frequency and with the position inside the gap.
Negative resistance of amplifiers is obtained by positive feedback. It is very easy to obtain it, without using an additionnal op-amp, by bootstraping with adequate resistors the input of the amplifier with the voltage accross the sensing resistor (overcurrent precautions are mandatory).
I have a strange point of view about negative resistance amplifiers. From the amplifier input, there is positive feedback, yes. But viewed from the impedance of the loudspeaker, it can be considered a negative feedback : the current through it decreases as its impedance increases. Then I tend to consider Stahl system as a kind of motion feedback system, disagreeing with his comment who said it is not. This is not an important question.
By controling the impedance of the amplifier, you can much extend the possibilities of a given loudspeaker or adjust it more adequately to your needs. Because it deals with the essential parameters of loudspeakers, I think it is a more interesting area to investigate than the endless discussion about the sound of amplifiers and the supposed horrible effects of feedback.
~~~~~~ Forr
§§§
I agree that the AC-bass technique is a very nice piece of engineering. I work at the department where Ståhl made his master thesis, but that was before my time there. Recently there has been a few DIY projects on swedish audio fora, building ACE subwoofers. The guy I share office with has even tried a DSP implementation of ACE, which is pretty cool. The parameters (Mms, Cms etc) can be modified in real time with a slider on the computer screen.
Hi Svante
"The guy I share office with has even tried a DSP implementation of ACE, which is pretty cool. The parameters (Mms, Cms etc) can be modified in real time with a slider on the computer screen."
This could be a very useful tool to study psychoacoustics : for example, with a closed box, what are the just noticeable difference for Qtb, resonance frequency fb, etc..
I just had a look at your simulation tools. Great !
~~~~~~~~ Forr
§§§
"The guy I share office with has even tried a DSP implementation of ACE, which is pretty cool. The parameters (Mms, Cms etc) can be modified in real time with a slider on the computer screen."
This could be a very useful tool to study psychoacoustics : for example, with a closed box, what are the just noticeable difference for Qtb, resonance frequency fb, etc..
I just had a look at your simulation tools. Great !
~~~~~~~~ Forr
§§§
phase_accurate said:
ahh http://www.freepatentsonline.com/4118600.html !
Its not actually in the AES papers that I can see,we have 1978-1997 at UNI,I checked 78 and 79,not there in the articles anyway. Is that where I find the patent number?
Thats quite cool,it would certainly have its uses.
"Synthesis of loudspeaker mechancical parameters by electrical means : a new method for controlling low-frequency loudspeaker behaviour"
by Karl Erik Stahl
The paper was presented at the 61st Convention of the AES, 1978 November 3-6; revised 1981 April 9.
Patents "ACE-bass method" :
USA 4,118,600
Sweden 7603585-6
The paper can be found in the AES anthology, volume 2, "Loudpseakers", pages 241 to 250.
Another interesting AES paper, explicitly based on Stahl's ideas :
"Extended low-frequency performance of existing loudspeaker systems" by R. Normandin which can be read in the AES Anthology, volume 3, "Loudspeakers" p61-65. The final embodiment needs two op-amps, a power amp, less than ten resistors and one capacitor.
Yamaha has patented a method where almost any kind of power ampilfier can be used, for example, one which the user already owns.
One of the most fascinating aspects of Stahl's ideas is that it can be applied to any kind of load : closed, bass-reflex, transmision line, etc...
~~~~~~~~ Forr
§§§
by Karl Erik Stahl
The paper was presented at the 61st Convention of the AES, 1978 November 3-6; revised 1981 April 9.
Patents "ACE-bass method" :
USA 4,118,600
Sweden 7603585-6
The paper can be found in the AES anthology, volume 2, "Loudpseakers", pages 241 to 250.
Another interesting AES paper, explicitly based on Stahl's ideas :
"Extended low-frequency performance of existing loudspeaker systems" by R. Normandin which can be read in the AES Anthology, volume 3, "Loudspeakers" p61-65. The final embodiment needs two op-amps, a power amp, less than ten resistors and one capacitor.
Yamaha has patented a method where almost any kind of power ampilfier can be used, for example, one which the user already owns.
One of the most fascinating aspects of Stahl's ideas is that it can be applied to any kind of load : closed, bass-reflex, transmision line, etc...
~~~~~~~~ Forr
§§§
GI
>Is it available? Can it be DIY'd?
Yes.
See US patent n° 5,206,912 by Masao Noro
at Patents Free On Line.
It consists in the inclusion of an op-amp ahead of the power amp. with an overall feedback about 26 dB around the whole and a local feedback of around 20 dB (network applied to the - input made a 10 kOhm and 1 kOhm) and around the op-amp. In parallel with the 10 kOhm resistor, there is a capacitor of 6 nF which limits the open loop bandwidth to 2.65 kHz -3 dB/o to maintain stability As the system is only going to work at low very low frequencies only, this is sufficient and most power amps should work ok.
The patent is not limited to this clever tip.
As always when dealing with negative resistance, include overcurrent protection at the output of your power amp. A polyswitch or two 12V car lamps were recommanded by Ian Hegglun for such experiences ("Speaker Feedback", Electronics World, may 1996, p378-382) to protect the voice-coil of your loudspeaker to be burnt. I suggest you start with low cost units to see how spectacular the results can be.
~~~~~~~~ Forr
§§§
>Is it available? Can it be DIY'd?
Yes.
See US patent n° 5,206,912 by Masao Noro
at Patents Free On Line.
It consists in the inclusion of an op-amp ahead of the power amp. with an overall feedback about 26 dB around the whole and a local feedback of around 20 dB (network applied to the - input made a 10 kOhm and 1 kOhm) and around the op-amp. In parallel with the 10 kOhm resistor, there is a capacitor of 6 nF which limits the open loop bandwidth to 2.65 kHz -3 dB/o to maintain stability As the system is only going to work at low very low frequencies only, this is sufficient and most power amps should work ok.
The patent is not limited to this clever tip.
As always when dealing with negative resistance, include overcurrent protection at the output of your power amp. A polyswitch or two 12V car lamps were recommanded by Ian Hegglun for such experiences ("Speaker Feedback", Electronics World, may 1996, p378-382) to protect the voice-coil of your loudspeaker to be burnt. I suggest you start with low cost units to see how spectacular the results can be.
~~~~~~~~ Forr
§§§
As Hegglunds article points out this idea has been re-invented every 10 years from the golden age of hi-fi in the early 1950's
I am surprised that nobody has mentioned the Audiopro subwoofers that were the result of Stahl's work. Stahl took the negative resistance idea a stage further and electronically tweaked the compliance and moving mass.
Also surprised that nobody has mentioned the key reason for using negative resistance, that is to match the Qts (more correctly the Qes) of a particular drive unit you like to a cabinet size that you might have already.
Rod Elliott has an article on this:
http://sound.westhost.com/articles/cscaling.htm
but he does not mention that you can reduce the Qts of a driver by negative resistance. This has the beneficial effect of damping the bass response
I am surprised that nobody has mentioned the Audiopro subwoofers that were the result of Stahl's work. Stahl took the negative resistance idea a stage further and electronically tweaked the compliance and moving mass.
Also surprised that nobody has mentioned the key reason for using negative resistance, that is to match the Qts (more correctly the Qes) of a particular drive unit you like to a cabinet size that you might have already.
Rod Elliott has an article on this:
http://sound.westhost.com/articles/cscaling.htm
but he does not mention that you can reduce the Qts of a driver by negative resistance. This has the beneficial effect of damping the bass response
consort_ee_um
Find below what Elliot has written about negative impedance amplifiers.
http://sound.westhost.com/z-effects.htm
He used to have a much larger article on the subject, which seems not to be anymore on his site, and was not keen on the technique. I think that he never undertood how to use it. Commercially they are at least two huge successes using it, Audio-Pro (sorry, not mentionned probably because not DIY) and Yamaha (sorry, mentionned because it's a well known company...) .
Note that positive feedback and negative resistance may probably beknown before negative feedback.
~~~~~~ Forr
§§§
Find below what Elliot has written about negative impedance amplifiers.
http://sound.westhost.com/z-effects.htm
He used to have a much larger article on the subject, which seems not to be anymore on his site, and was not keen on the technique. I think that he never undertood how to use it. Commercially they are at least two huge successes using it, Audio-Pro (sorry, not mentionned probably because not DIY) and Yamaha (sorry, mentionned because it's a well known company...) .
Note that positive feedback and negative resistance may probably beknown before negative feedback.
~~~~~~ Forr
§§§
Yes, Rod does not like the technique but it works for me
See http://www.diyaudio.com/forums/showthread.php?s=&threadid=19114&perpage=10&highlight=&pagenumber=1
Scroll down to Moamps posting for my humble contribution
See http://www.diyaudio.com/forums/showthread.php?s=&threadid=19114&perpage=10&highlight=&pagenumber=1
Scroll down to Moamps posting for my humble contribution
I once built such a thingie. The result was quite astaounding. With today's wide availability of specialized subwoofer drivers I wouldn't use it anymore.
AFAIK not even Sonab (the owner of the trademark AudioPro) uses it anymore.
I have both articles (i.e. Stahl and Normandin) on photocopy and I copied them from the Journal ond not convention preprints. I can have a look when they appeared if someone is interested.
Regards
Charles
AFAIK not even Sonab (the owner of the trademark AudioPro) uses it anymore.
I have both articles (i.e. Stahl and Normandin) on photocopy and I copied them from the Journal ond not convention preprints. I can have a look when they appeared if someone is interested.
Regards
Charles