We already have current drive, it's current that drives the drivers, so to speak. Speaker designed with series crossovers will work with High Z current drive. but there is of course issues/problems as the impedance peaks of mechanical resonances sets great demands on the voltage swing capabilities of the driving amplifier.
If you design drivers from the bottom (like I tend to do) then theres the option of making a low Z driver. like with an impedance of maybe 1/2 to 1 ohm, this will cure many of the base resonance problems. You could also mount the drivers in a not so air tight sealed box (a way to smoothe impedance peaks in the base region)
I think if you really look closely, then current drive could be very exciting from a performance point of view, but as it's a paradigm shift I have big doubts of how to make it commercially fly.
If you design drivers from the bottom (like I tend to do) then theres the option of making a low Z driver. like with an impedance of maybe 1/2 to 1 ohm, this will cure many of the base resonance problems. You could also mount the drivers in a not so air tight sealed box (a way to smoothe impedance peaks in the base region)
I think if you really look closely, then current drive could be very exciting from a performance point of view, but as it's a paradigm shift I have big doubts of how to make it commercially fly.
whaattt!
This is crazy !
Something can not be JUST current driven.
Wherever there is a current there must be a voltage.
Unless you have a loudspeaker with 0 R.
Now , if you took a dual VC sub and fed back the 2nd coil's EMF
back to the amp , you might have something there .... but (I) is never
alone.
OS
This is crazy !
Something can not be JUST current driven.
Wherever there is a current there must be a voltage.
Unless you have a loudspeaker with 0 R.
Now , if you took a dual VC sub and fed back the 2nd coil's EMF
back to the amp , you might have something there .... but (I) is never
alone.
OS
I summarize my thought on this in http://www.diyaudio.com/forums/solid-state/240712-cfa-topology-audio-amplifiers-434.html
Rather surprised that my Powered Integrated Super Stuff STILL represents the SOTA today as the work I did on this was late 70's, 80's & 90's.
The advantages of current drive are well summarized by Mills & Hawkesford. But there is no need to adopt their extra windings or even their Transconductance Power Amplifier Systems for Current-Driven Loudspeakers to get the benefits.
Erik Stahl's method http://www.aes.org/e-lib/browse.cfm?elib=3890 is far more elegant and powerful.
The problem still remains that no one knows how to design current drive speakers for good performance .. unless you believe SET amps & full frequency range moving coil units are the bee's knees
But there are many advantages for subs as the amp has to be integrated with the speaker & box anyway.
___________________
There's a lot of liquid BS in the EDN article and its comments. But dis beach bum can't affort to buy his book so waits with interest his part 2. I'm not hopeful though.
The definitive work on speakers as microphones is Loudspeakers as High-Quality Microphones by Great Guru Baxandall of course. Important subtexts on Reciprocity & Transformer Design.
___________________
LdSpkrPro, do you work for any commercial speaker companies?
Rather surprised that my Powered Integrated Super Stuff STILL represents the SOTA today as the work I did on this was late 70's, 80's & 90's.
The advantages of current drive are well summarized by Mills & Hawkesford. But there is no need to adopt their extra windings or even their Transconductance Power Amplifier Systems for Current-Driven Loudspeakers to get the benefits.
Erik Stahl's method http://www.aes.org/e-lib/browse.cfm?elib=3890 is far more elegant and powerful.
The problem still remains that no one knows how to design current drive speakers for good performance .. unless you believe SET amps & full frequency range moving coil units are the bee's knees
But there are many advantages for subs as the amp has to be integrated with the speaker & box anyway.
___________________
There's a lot of liquid BS in the EDN article and its comments. But dis beach bum can't affort to buy his book so waits with interest his part 2. I'm not hopeful though.
The definitive work on speakers as microphones is Loudspeakers as High-Quality Microphones by Great Guru Baxandall of course. Important subtexts on Reciprocity & Transformer Design.
___________________
LdSpkrPro, do you work for any commercial speaker companies?
Last edited:
As far as I remember there was even a thread here by the author. He was blasted mainly for his very abrasive way he said something to the effect that all speaker and amp designers up to now were morons and he was their savior.
That didn't go down well, and is also the reason that I closed his book after the first chapter in disgust.
Possibly he has interesting things to say, but boy does he say it badly!
C'est le ton qui fait la musique...
Jan
That didn't go down well, and is also the reason that I closed his book after the first chapter in disgust.
Possibly he has interesting things to say, but boy does he say it badly!
C'est le ton qui fait la musique...
Jan
Last edited:
This was the thread which I've laboriously ploughed through. http://www.diyaudio.com/forums/vendors-bazaar/157787-secret-tube-amplifiers-revealed-much-more.html
Alas, its practically all "yus guys are all idiots & deaf" with several true masters at this chipping in. I don't think any speaker designer thought it worthwhile to participate.
So I didn't find any tips on designing speakers for current drive and Part 2 of the EDN article was similarly full of obfuscating stuff which had no bearing on the advantages of Current Drive.
The Amazon preview of the book has also disappeared.
I is never alone and will have to be associated the a variable V. But regardless how you look at it, it is the I that makes the cone move in the magnet field. Speakers are current driven, but traditionally controlled by voltage.
Now over the years V(out) amplifiers has set the framework for all speaker developments. the opposite could just as well have been the case.
I am quite confident that current drive could produce just as good results as voltage drive, one just have to consider how to deal with a totally different approach to driver design and system integration, filtering etc. etc.
Now over the years V(out) amplifiers has set the framework for all speaker developments. the opposite could just as well have been the case.
I am quite confident that current drive could produce just as good results as voltage drive, one just have to consider how to deal with a totally different approach to driver design and system integration, filtering etc. etc.
Then you add the contributions of another 'outside the box' thinker called Eduardo de Lima and the theory/ideas behind his AudioPax amplifiers - something different.
Yes, Esa Merilainen's book isn't an 'easy read' at all and obviously not everyone's 'cup of tea' - perhaps it was written to get greater attention, or create discussion/controversy - we're still talking about it, so it did stay in our minds.
Rod Elliot (esp site) has added some words about a similar related project, too.
Yes, Esa Merilainen's book isn't an 'easy read' at all and obviously not everyone's 'cup of tea' - perhaps it was written to get greater attention, or create discussion/controversy - we're still talking about it, so it did stay in our minds.
Rod Elliot (esp site) has added some words about a similar related project, too.
As far as I know, the very clever Erik Stahl's method deals only with the bass region. Initialy, before the special positive/negative feedback to control the output impedance is applied, the power amp is supposed to be a voltage source.
Many advantages of current drive concern the whole audio band, even the bass region where it should decrease the harmonic distortion of the driver; at the well known expense of the lack of electrical damping of the main resonance which is the only great difficulty the topology has to overcome.
I did not bother Merilainen's style. Sure, english is not his natural language. Technically, I enjoyed his book.
It is worth mentioning that, according to Merilainen, the advantages he sees in current drive become to appear with modest resistance drive, > 8 Ohm.
No I is never alone, neither is V
But the definition is clear:
Current drive means that the current is controlled and does not depend on the load impedance;
Voltage drive means that the voltage is controlled and is independent of the load impedance.
Of course, there are limits - if the impedance is infinite, the current will not be maintained; if the impedance is zero, the voltage will not be maintained.
But it's not a matter of 'always I and V go together' - they do, but the difference between voltage and current drive is not ambiguous.
Jan
The 'advantages' of current drive are also best seen in an active speaker with integrated amp.
The CONs of current drive go far beyond what Merilainen is prepared to reveal. In the http://www.diyaudio.com/forums/vendors-bazaar/157787-secret-tube-amplifiers-revealed-much-more.html thread, many pseudo gurus pontificate on F=Bli etc but NONE of them complete the analysis and show why Voltage Driven conventional drivers are predominant today. ie none of them are real speaker designers and understand how moving coil speakers work.
There are very good reasons why Rice & Kellog's invention is the BEST solution .. from the cheapest to the most expensive and elaborate passive speakers .. and will remain so for the forseable future.
That's not to say there are no advantages for Current Drive but the important ones are given by Stahl with better consistency & reliability.
Today however, zillion point DSP allows us to adopt a 'pure' Current Drive approach. Its the ..
A brute force zillion point approach possible today is measure accurately speaker Z (not that easy) and tailor the output of a High Z amp to suit. This would give some but not all the advantages of the above systems cos it wouldn't "know" what the speaker is up to.
... from my http://www.diyaudio.com/forums/solid-state/240712-cfa-topology-audio-amplifiers-434.html post .. though that has some serious disadvantages too ..
In case it isn't obvious, I've tried all the methods in that post including a crude version of the DSP in 'real life' circa 1990.
Last edited:
As I said somewhere -- the biggest issue is the harmonices generated by the load get thru the passive XO for Mid and Treble etc if the low Z from the amp output isnt right close at the speaker XO; To some extent the cable Ls acts like a freq dependant buffer between amp and XO/SPKRs. high Ls speaker interfacing cables should be avoided.
True?
Thx-RNmarsh
True?
Thx-RNmarsh
Last edited:
If it can be figured out how to design class A SE amp with current drive and maybe with two or three FETs and without global feedback, I'm in
FIRST WATT F2
?
Would you mind to explain where is the non-linearity to create harmonics? Or was it a poor unstable amplifier?
This scope shot is of a high power speaker being driven by a 300 watt amp at Fs. The orange signal is amplifier voltage. The blue signal is the voltage across a shunt in series with the speaker. The current waveform is very sinusoidal up to about 2 amps peak. Above that, the current waveform folds back on itself and just gets more and more distorted. The peak current never becomes larger even though the voltage increases.
I am at a bit of a loss to understand what is meant by amplifier damping. What is the amplifier damping? It is certainly not damping the back emf. It is shorting the back emf and causing a significant distortion of current.
I am at a bit of a loss to understand what is meant by amplifier damping. What is the amplifier damping? It is certainly not damping the back emf. It is shorting the back emf and causing a significant distortion of current.
I modified the setup to include an op amp in front of the same power amp. The op amp negative input is sampling the voltage on the current shunt. Attached is a capture of this setup with the same speaker at max voltage. You can see that the peak current is allowed to go much higher in current drive mode than it can in voltage drive mode.
This setup is unstable. Just my non-skills at op amp design. I plan to have a more stable setup to repeat this test in late March. When I do have it running, I will measure distortion of the acoustic output for the voltage drive vs the current drive. Until then, I only have speculation to share.
My guess is that the current drive is going to have measurably lower acoustic output distortion.
This setup is unstable. Just my non-skills at op amp design. I plan to have a more stable setup to repeat this test in late March. When I do have it running, I will measure distortion of the acoustic output for the voltage drive vs the current drive. Until then, I only have speculation to share.
My guess is that the current drive is going to have measurably lower acoustic output distortion.
Last edited:
You're driving the speaker at a very high power and seeing the back EMF reduced to under 2%. The amp is doing its job- that's what damping is, the shorting of the back EMF. If you run the speaker near its limits, you expect the current to be "distorted" for reasons that a few minutes of Klippel analysis will make clear.
I do understand. The way I look at it, the amp is sampling speaker voltage and corrects voltage to be perfect. The back emf distortion shows up in the speaker current. That is the only place the distortion can show up since the amp feedback is correcting voltage. I replicated this behavior with high power current pulses. I created current pulses and saw the voltage distortion.
The current drive amp is correcting current. The back emf distortion shows up in the voltage. The back emf is playing into an open circuit so the back emf causes no current distortion.
My assertion is that since the speaker is driven by current, the current drive (perfect current waveform) will result in less acoustic distortion. I am willing to be proven wrong. Its just an assumption.
To test the difference of distortions at a given frequency between the two ways of driving a loudspeaker discussed here, the voltages across it have to be identical at that frequency so that the acoustic levels will be the same. A 4 Vrms may be a good value and won't introduce significant thermal effects.
In voltage drive when the voice coil moves away from its central position, the back-EMF is reduced due to the lower (Bl) factor hence more current is drawn from the amplifier (cf Martin Colloms "High Performance Loudspeakers" Wiley, 5th edition, page 91, 'Instability').
This phenomenon introduces non linear distortion and does not happen in current drive.