Mr. Esa Meriläinen,
first read about the concept – I was shocked.
The right hand rule – says it all – it is the CURRENT that drives the coil of my speakers, and not the voltage applied to it.
I learned about this in the eighth grade of elementary school, during
some very basic lessons of physics. Quite a few of us, audiophiles, actually did. Simply for the sake
of passing the exams.
Not excluding the whole audio industry.
Your book has the potential of shaking the very foundations of the whole audio industry as we know it... Yet it seems that you are from a totally different industry. How did this bridging to the audio domain actually take place?
I was perhaps about ten when I began to pay attention to remains of a tube radio in our attic. Pretty soon I dismounted the speaker from it and began to wonder its construction. How could it be that that kind of framed paper disc was able to emit lifelike sound? I began experimenting by pushing the two wire ends in the various connectors in my brother's portable cassette radio. The excitement was great when I finally hit the right holes and the speaker began to play, even louder than the radio itself.
By 17, I had constructed perhaps a dozen audio amplifiers for various purposes and a heap of other electronic gadgets from books. Already then I become aware of certain sonic differences between some amps but of course didn't have the qualification to explain them.
In my student years in Tampere I had a job that involved assessing circuit topologies for analog computation, especially multiplication. This also made me acquainted with the principles of handling signals as currents. After one such working day, as I was resting at home with my three-way boxes in sight, the idea entered my mind: "What if...?" The next thought then was: "Actually, why not?" I immediately realized that the idea made very much sense, and I planned to one day take the matter into closer consideration.
It was, however, only a couple of years later when I finally took up the issue and made some first experiments with current-drive. I didn't expect any wonders, but even from the first cadences I was left frozen and jaw-dropped. It was a night-and-day difference for the better compared with anything I had heard previously anywhere. From that moment on, I became really interested and started to find out more about the subject and apply my knowledge on it.
Your background is academic, yes? Can you tell me something about what you do professionally? Is there any "link" to the audio domain, or is it more of some kind of hobby?
I have an M.Sc. degree in electronics engineering, but my job occupations are not related to audio, as there are not around companies who are willing and able to focus on current-driving. I can also speak more independently and credibly when there are no affiliations behind. Having said that, my career is not of much meaning here, nor have I made it public anywhere.
I am not a passionate audiophile in the sense that I would feel a constant need to upgrade or tune my components or that I would have shelves full of discs. Rather, it is this one fundamental issue and also some grave misconceptions related to it that I am striving to make known.
As you probably know... I am already a converted "Believer". But how could you, in few short words, explain, or at least provide an outline, of the "proof of concept"? Why "Current" and why is it a "better" way to go?
An electrodynamic transducer is a current-controlled device that in the end is faithful only to current according to the well-known, unmistakable law F = Bli. To depart from this plain directive in an accuracy-demanding application would require strong, scientifically valid justifications, that, however, nobody has ever presented. Thus, the burden of proof actually lies on those who choose to disregard the above rule.
In short, it is the critical voltage-to-current conversion that becomes severely corrupted under voltage drive as a result of the manifoldly uncontrollable electromotive forces (motional and inductive) induced in the voice coil. It can be easily demonstrated that this VI conversion, when left as the duty of the speaker driver, is indeed the most significant source of nonlinearity distortions throughout the mid-frequency region.
So, it seems that all of the high-end audio amplifiers should be current output amplifiers. Why has the whole industry gone "astray"? How is it possible that they got it "all wrong"?
I would not even limit current-drive to so-called high-end products. Ordinary listeners would quite as well benefit from the non-messed-up operation of their speakers. Virtually all mankind is affected by this misconduct.
Actually I am not suggesting that all audio amplifiers should work in pure current mode. Subwoofers can do quite well also as they are, as the flaws of voltage drive are mostly effective above the bass frequencies.
I think there is the ageless problem of human consciousness: to know the truth and still go with the lie. The fields of science and technology just don't seem to make exceptions.
From an engineering perspective, despite a couple of reported experiments, there has been quite nobody to characterize and speak out the real detriments of voltage drive. In the early days of audio, in turn, development may have been driven more by cheapness, simplicity and consistency with available technology rather than by concerns with current purity. We also have to recognize the quality of the storage and broadcast media of those times and the fact that there were no spectrum analyzers or computing tools available.
What about the speakers? Is it not that present technology speakers are "optimized" for the voltage output base amplifiers? Do I need "special" current output compatible speakers to achieve the best possible results?
Speakers naturally have to be designed to work with the concept. Driver requirements are also somewhat different. Applying pure current-drive for existing voltage speakers is not so advisable due to frequency response alteration.
From my initial experiments it seems, that the standard speaker sets fair very nicely, but is there any potential problem with using "standard speakers"? If indeed – then is there any simple way by which I could "adapt" such standard speakers, so as to optimize them for current drive mode of operation?
With "standard speakers", the frequency response will be re-shaped according to the impedance curve which typically means peaking in the bass and crossover regions. With suitable EQ tools, however, such adaptation may be possible, but this doesn't yet substitute for a dedicated current design. In general, the potential of current-drive should never be judged simply by feeding current to an existing voltage speaker, only to note the boomy bass, as too often is seen done.
Or maybe it is a question of cost? Maybe it is so that such current driving amplifiers are more complex to construct, and therefore more costly to produce?
I have built quite a few without encountering any special difficulties in that. Mostly only the feedback arrangement is different, and in the basic configuration the speaker is not tied to the ground. Distortion and other characteristics are very similar to those achieved in the voltage world.
The book that you wrote about this fascinating topic – Are these experiments a part of your professional activity, or did you perform them strictly so as to build an "experimental" verification and a foundation for the book?
All has been a professionally qualified, devoted, private hobby, no other parties involved. The experiments have been performed and the circuit ideas devised in the course of many years for the sole purpose of this project. Ideas have also continued to flow and evidence grow after releasing the book.
Do you find time (or "make" time) to listen to good music? What does your personal "stereo system" look like? Do you personally, for your own listening pleasure, "current-drive" your speakers? Do you Walk the Walk so as to Talk the Talk? Did you have to "modify" those speakers?
Actually I can't find time to listen as much as I would like. I prefer acoustic recordings that offer some sort of beauty. Anything with a syncopated beat is not my way. My personal stereo setup still consists of those projects I have presented, with a Sony SACD player as the main source. New projects may come, but not a slightest longing for the "damping factor" school.
Is there "anybody out there"? I mean, some company, some producer of audio equipment, of whom you know that they actually implement, use the current drive philosophy and apply such a topology into their commercial equipment?
Not quite anything that I could recommend as a whole. There have been some amplifiers around that provide a special knob to "match" the amp with the speaker; which is quite obviously only an adjustment of the amplifier's output impedance within a given range. These can be helpful in experimenting though are not yet current-drive.
When a current output mode amplifier suddenly "looses" contact with the load... what happens then? Will the voltage of the output soar sky high up to theoretical infinity? Is it "safe" to rip out the speaker cable from a current driven amplifier, while the device is operational?
The output voltage hits its limits as determined by the supply voltages, but this usually doesn't pose a danger when there are protection diodes in place. It has happened to me countless times without damages. A short between the speaker wires is also much less a problem in a current-output amp than it is in traditional amps.
Is it actually true that the output impedance of such a... modulated current source... if I may call it – is theoretically close to infinity? Sort of hard to imagine for the average audio consumer ...
In an ideal current source, the output impedance is infinite. For practical power amplifiers, it is enough that the impedance is very high compared to the load impedance. We may define a "current-drive index" as the ratio of the output impedance to load impedance and use it as a measure of ideality in an analogous manner to the "damping factor" used for voltage amps.
I would expect that "forcing" a current against the "mechanical" will of the speaker, of the air pressure, of the elasticity of the suspension, all the possible phase shifts... could be perceived as a potential source of problems. If I "apply" a voltage – I sort of say: OK, Dear Speaker, so here you have a voltage and please adapt to it to the best of your capability ... Take your time and I extravagantly allow you to adapt – with all your lag and all your phase shifts, to such a new "situation". But applying a brutal force CURRENT – instantaneously – sort of "forcing" the current to FLOW through the speaker, to flow AGAINST its WILL – ( the "will" – being the back electromagnetic force that is "generated" by the coils moving within those magnetic fields, and many other things) – this seems that it can be a source of potential problems. Maybe some nasty voltage spikes... or something? Or maybe – even instability or oscillations? Is there any problem that might arise out of the fact that we are "forcing" the current to flow? Instantaneously?
The amplifier (voltage or current mode) always sees the speaker as a certain load impedance and nothing else. That is, the effects of EMF are already included in this impedance, and there are no such EMF effects that we would need to consider or handle separately or in the time domain. Thus, the amplifier really knows nothing about EMF or the various effects that shaped the impedance. It only sees the resulting electrical impedance and acts on it.
When we drive any impedance by voltage, the resulting current becomes exactly determined by the impedance, but we can just as well drive the impedance by current in which case the voltage becomes correspondingly determined by the impedance. In principle, there is not any difference in the difficulty of driving; it's just the same linear system turned around.
Please tell me about the audience, for which this book is targeted. Who is the potential "reader"? Will a part time audiophile, a consumer, who loves good sound and good music, but is not deep into electronics and stuff – will he have a chance of understanding what you are writing about? On the other side: If we are an electronics guru, a person who designs amplifiers... will such a reader find your book too elementary?
It has been
written to be useful for varying levels of audiences, and no prior expertise in
the field is necessarily required. The symbology used is as simple as possible,
not referring to the general, difficultly configured symbology of the art.
Educated readers can assimilate the whole content readily; those who are not so educated but willing to discover the principles of analogue linear systems are provided a tutorial in the appendixes. However, even without knowledge in math, the message and instruction can be understood and applied.
Correct if I am wrong, but this is as yet your only one piece of writing on the subject, or are there other items of literature, that you wrote about this? Or if not, would you be kind enough too hint on some extra reading, just to get a better understanding of the whole concept? Will you be writing "more" on these topics?
My website at www.current-drive.info provides some first reading.
A review of the book with a short description of each chapter can be found at Inner Audio.
Recently, the EDN network has published my article about the flaws of voltage drive. (As of today, the second part is yet being processed.)
Hifi & Musik, a Swedish paper magazine, published in their October 2013 issue a four-page article on current-drive based on the book and interviewing me by mail. The heading was "Current, not voltage!". I am also offering a paper about my distortion comparisons to the AES journal, but their stance is yet to be seen.