Joseph K said:
Well, here we go...
You are perfectly right in this observation, but it holds for full range lone drivers. Admittedly I did not stress that for multi-way speaker systems, displacement nonlinearity is significant mostly for the low frequency driver.
The remark about acceleration was with reference to previous posts where experimental work was cited, linking acceleration with sound pressure, implying force.
With respect to Hawksford's paper, I remarked then I was unconvinced and wanted to think and read some more. Reasons behind is frankly I am uneasy about the velocity-current (eqn. 2) relationship which derives from the model of fig. 2 (a), where the motor transformer converts current to velocity and voltage to force respectively.
As said I am not implying it to be wrong, just uneasy. If we consider as a bare starting point a freely moving resistance-less coil within the magnetic gap, then for an impossed current we get the Lorentz force proportional to current, while when it is moving, we measure an induced emf proportional to velocity.
The situation gets progressively more complex when we add mass, suspension, cone and air coupling, resistance etc. Yet I fail to grasp how it comes cause variables somehow switch places for effect results.
Will keep pondering this as time allows (which is not plentifull).
Thanks for the reply George!!
Rodolfo
Hi Graham,
I did not try this yet. However note that the LM3886 application note (and maybe other chips) use a 0.22 ohms in the speaker line. In one simulation I noticed that it stops ringing when using square waves . I have a question here.
Some listeners have found single core 'thin' speaker wire better sounding than fat cable. Could it be that the cable has high enough dc R to stop ringing and hence sound better ? This of course would vary with the system being used I guess.
Cheers.
Hi Paolo,
I have little time and have not a setup to "decently" analyse the THD of the spks.
This does of me a frustrated person...
Ciao
Mauro
I have little time and have not a setup to "decently" analyse the THD of the spks.
This does of me a frustrated person...
Ciao
Mauro
Graham Maynard said:
I must admit that it really sounded (looked) like gibberisch. With the additional explanation, it becomes clearer. But I still don't understand the term "resist its return". Isn't the amp output just short-circuiting the EMF voltage that manages to get through the xover to its (the amps) terminals?
Why are you always trying to cram long strings of terms into seemingly endless statements that are so difficult to decipher? It is difficult enough without the language barrier that must be formidable for many here. Don't you WANT us to understand you?
Jan Didden
janneman said:
Everyone does, niether you nor I ever wanted to trap in the "misunderstanding criteria of human psychology"...
cheers,
K a n w a r
Ha! Just try listening to some politicians sometime.😀 There are many who deliberately use uncommon or just plain made up words along with convoluted grammar specially designed to sound impressive without actually saying anything useful at all to cause misunderstanding, for a variety of reasons.Workhorse said:
Graham Maynard said:
Hello, Graham
You are referring to standing waves, due to mismatched impedances at the sending and receiving end of a transmission line, right?
It's an interesting property of transmission lines that they have an characteristic impedance, and there will be reflections both if the driving and terminating impedances (the ones placed at the driving side and receiving side of a line) are higher or lower than the line impedance.
Now, it's important to note that in a real line, every time the electrical energy flows throug it there will be a loss, and all energy will finally dissipate.
A twisted wire line has an characteristic impedance of about 100 ohms, some very high end cables have lower values.
Now, let's see:
Driving side:
A very high damping amp presents to the line an audio frequency driving impedance of under 0.8 ohms; a low damping amp (like an SE tube) more like a couple of ohms.
However, we are talking VERY high frequency artifacts over here, and both amps will present to the line probably an capacitive impedance at the really important frequencies, and the line will not be terminated at the driving side, no matter what type of amp one uses.
Terminating side:
The load may be a simple speaker or a speaker/Xover combo, but both with an audio frequency impedance around 8 ohms, generaing reflections.
With a simple speaker, impedance rises to a point, there will possibly be an out of band high frequency ressonance (due to the distributed capacity of the coil winding), terminaton may be capacitive and there will be reflections.
With speaker/Xover systems, it's a harder guess, but all evidence points to reflections also.
Placing a resistor as a damping element in a line may not work - if, at the VERY high frequency envolved the resistor is inductive, it will be useless.Besides, it value would have to be such taht combined with the real amp/speaker's impedance it would terminate the line in resistence AND phase.
This is anyone's guess.
And since I've made references to VERY high frequencies, let's see:
The propagation speed of a typical isolated copper pair or coaxial cable is of the order of 200,000,000 (200^6) meter/sec.
this means that for a 3m typical lenght in only 3/200^6 = 15 nanoseconds the signal will reach the end of the line and will be reflected back.
This means the frequency we are talking about is 66.66.. MEGAHERTZ...
There are evidences to believe that such a frequency is not audible.
And then what goes on?
The shortest time span between 2 audio signal may be taken as 1/20 KHz or
50 microseconds.
During this time, the line reflection has travelled back and forth in the line 50^-6/15^-9= 3333 times, and at each time it will loose energy (all kind of losses one can concieve), and will be subdued beyond any possible identification, by whatever means - including golden ears.
And BTW, the 0.22 ohms JLH uses on the output of his anps is a very clever way of avoiding the classical inductor to warrant amplifier stability aganist oscillation - it has nothing to do with avoiding line reflections (i've read his original article in Wireless World).
Sorry, your pet theory doesn't apply in the real world.
Jorge
ChocoHolic said:
Hi,
if I remember right a CFB amplifier will control the cone movement very firmly, if you push on a loudspeaker cone connected to a transimpidance amplifier it will keep the cone very firmly at it's position, that will not happen in a VFB amlpifier.
This is happening because the CFB sense a (if I may use the very hot word) EMF current.
Hi all,
so are we wiser now regarding the "back EMF" from loudspeaker into an amplifier?
First an EMF is an inductive phenomena, google around and theres NOTHING about capacitive EMF(I just say this in case because I have not seen any one discussing closer what it really means and where it comes from/where this shortness is used).
EMF is a shortness I learned for many years ago in school meaning an opposing current from a component/device like an electric motor when you apply a voltage, if we look at loudspeaker elements, they are too motors (which have already been stated couple of times latest time here on DIYaudio).
When above is said we can start to look at the element and where it have it's capacitive and inductive part, for the highest frequencies (as a worst case reality) we can look at a tweeter.
A tweeter has it's highest capacitive part just above the resonance frequency which often is around 1 kHz reaching up to 3-5 kHz were it starts to be resisitve, not any high frequency for almost any decent amplifier.
Then there's the inductive part which is after the capacitive area with rising frequency(and also below the F0).
So is the inductive EMF any real problem for an amplifier, I would say no because the current is LAGGING and thereby it's a change happening way long after the voltage signal is seen on the output and the FB node normaly reacts within a much smaller timedelay to changes happening at the output.
If the amplifier can't handle a lagging current at 20 kHz I would say that amplifier has a FB bandwith much less then 20 kHz and extremely slow such, but as we all realize that can't just be the case with any good amplifier.
GM have mentioned a couple of times that the low frequency back EMF would affect on high frequency signal, that depends wethere the high frequency is comming from THE SAME element or not!
If the high frequency signal is above the crossover filter frequency then the signal will be distributed to the next element taking care of high frequencies and we wont have any modulation here.
But we could have a case where both high and low frequncy signals are emitted from the same element, yes it's true that loudspeaker elements are not very linear when they start work at high excursions, and of course the high frequency signal will start to be modulated by the low frequency signal IF!!! the low frequency signal is of such a scale it's start to drive the element into an unlinear area, this has NOTHING to do with back EMF problems and the modulation is happening solely IN THE element!
So what can happen in an amplifier subject to back EMF, I will state it very clearly theres only TWO paths:
1: is what the feedback node sees (if we use a feedback topology connected to the output).
2: the refflecting current going backwards through the output transistors, driver transistors all the way to the VAS which use to be such a high impedive circuit it wont reflect much backwards anymore.
First the feedback node, I would strongly say that the feedback node DOES NOT KNOW what kind of phenomena is affecting on the output voltage signal, so hopefully no one think that back EMF is doing anything "special" to the feedback node, it's a higly passive circuit in that matter and does what it has to do, CORRECT the output signal respectless whetere it is affected by unlinear and/or reactive load, or the forward unlinearity of the amplifier it self
The refflecting current, this is the only path a so called "external unlinear and/or reactive" load can make any influence to the amplifier at all IF it is an amplifier which use BJT's as output and driver transistors in EF mode, as soon as we have anythin else like a FET it will isolate to a certain degree the refflecting current opposing at the output.
Conclusion is in my oppionion there's no real problem with any decent amplifier design and reactive loads.
Finally I would not use the word EMF, just talking about reactive loads, and I hope everyone here would do the same, that is much more correct and involves both capacitive and inductive components which EMF dont do!
And thereby this discussion should be over! 😎
Regards Michael
A small note on feedback,
GM have one point when talking about Miller compensation, but this applies not only to "EMF" kind of distorted signals seen on the output.
The hint is: build "more stable" amplifiers! 😉
Michael
GM have one point when talking about Miller compensation, but this applies not only to "EMF" kind of distorted signals seen on the output.
The hint is: build "more stable" amplifiers! 😉
Michael
Ultima Thule said:
Micheal,
I think this is not correct. The voltage amp will damp the cone movements by short-circuiting the emf-induced current. In the CFB case, there is not such a shorting effect, so the damping relies only on the mechanical damping and whatever paths there are for the emf-induced current through for instance the xover. The cone will need much more time to come to rest.
Jan Didden
Michael,
Jan is right, no damping from the part of a transconductance amplifier.
Like if it were not there.
Ciao, George
Jan is right, no damping from the part of a transconductance amplifier.
Like if it were not there.
Ciao, George
.... and that is one reason why current drive amps often seem to have more bass. The resonance on the low side of the woofer is less damped and appears to give more bass. Especially with wide-band speaker drivers, that don't go very low, current drive can give you some extra lf extension that make them useable. But that extra bass is not very well controlled and often boomy and not 'tight' .
Jan Didden
Jan Didden
Hi all,
Current driver, no electric damping, is true.
But the ability of control of the electric damping of the cone movement is limited on account of Re.
If we use a current driver with equalization of (freq.) response or with motional feedback , the control improves notably ( and even the IMD ). It is obvious that this type of system is not applicable to the commercials SPKs, at least that is not uses a DSP that calibrates the frequency response on all the range ( in theory it is not a lot of difficult, is enough a system of automatic calibration in the frequency domin...)
but this is an other history....
Ciao
Mauro
Current driver, no electric damping, is true.
But the ability of control of the electric damping of the cone movement is limited on account of Re.
If we use a current driver with equalization of (freq.) response or with motional feedback , the control improves notably ( and even the IMD ). It is obvious that this type of system is not applicable to the commercials SPKs, at least that is not uses a DSP that calibrates the frequency response on all the range ( in theory it is not a lot of difficult, is enough a system of automatic calibration in the frequency domin...)
but this is an other history....
Ciao
Mauro
I forgot,
I am experimenting the Graham tests ( reverse driven ) on my circuits, and the linearity of phase and of damping "is feels".
This it may be in league with the work on Back_EMF or only "generically " they reactive load , and even to the amelioration of the internal linearity of the circuits, but of some it works better on the general linearity with the analyses of ( reverse driven ) that with that traditional.
Knowing that the SPKs have powers of " produce " voltages...
Ciao
Mauro
I am experimenting the Graham tests ( reverse driven ) on my circuits, and the linearity of phase and of damping "is feels".
This it may be in league with the work on Back_EMF or only "generically " they reactive load , and even to the amelioration of the internal linearity of the circuits, but of some it works better on the general linearity with the analyses of ( reverse driven ) that with that traditional.
Knowing that the SPKs have powers of " produce " voltages...
Ciao
Mauro
mauropenasa said:
Mauro:
Did you conclude this from direct experience or based on the Hawksford paper?
With respect to the latter, I've been studying the issue and found distortion does not necessarily improve with current drive, but will round up and post on this in the next days.
Rodolfo
Hi, Rodolfo
I are much pleased about knowing your conclusions, and I appreciate your prudence.
It seems me bewilder that the "Hawksford theories" are like that harvest in discussion.
It comes by me lie an inverse question:
Which mathematical analysis show the advantages ( on THD and IMD ) of the voltage driven ? 🙄
Ciao
Mauro
I are much pleased about knowing your conclusions, and I appreciate your prudence.
It seems me bewilder that the "Hawksford theories" are like that harvest in discussion.
It comes by me lie an inverse question:
Which mathematical analysis show the advantages ( on THD and IMD ) of the voltage driven ? 🙄
Ciao
Mauro
mauropenasa said:
Mauro:
In a nutshell, for the assumption of idealized linear speakers, there is no difference for voltage vs. current drive except for effective Q where voltage drive is recognized as better even by Hawksford.
For the nonlinear situation, the real issue is to analyze what is formally called as sensitivity, that is, how much does affect the output a given variation in an (ideally constant) parameter such as Bl or Zm etc.
What I've found is that in this respect - despite the deceiving simplicity of eq. (2), in fact eq. (1) - voltage drive - displays slightly better immunity.
As said I will round up and post whatever I conclude in the next days as soon as I have some time to tidy up a decent document.
Rodolfo
Hi ashok,
I was not aware of the chip-amp recommendation.
Yes a thin cable can stop interface induced ringing, but sound better ? Only if you are very lucky !
When a thin wire is used, loudspeaker reproduction loses attack on transients and there is a tonal change.
If you use thin cable to a 2 or more way loudspeaker system that does not have near perfect impedance compensation it is likely to sound completely wrong, and not even the best graphic equaliser could compensate because the problem arises with waveform dynamics as well as amplitude.
I'd love to hear if that Creek sounds better with its own series 0.22 resistors.
Hi Jan,
Post#124. Hey I can't help what comes out when some question or a comment fires my neurones - we're all different.
I'm glad we can discuss this though, for I respect your views, but I am not capable of guessing what part of my understanding others think needs to be questioned !
Yes, that is my take too - a NFB loop controlled solid state amplifier is like a short circuit to the speaker.
I have already started going through this and saved it on my desktop, but the last couple of days leave me struggling for coherent thinking.
Hi Kanwar,
I cannot know or assume in advance what anyone else understands, especially when there are different ways of looking at fundamental theory.
So how can I be responsible for making someone understand, if they don't say what it is about my words, they doubt ?
Hi Jorge,
Re your first line question in Post#127. No ! Line reflections my pet theory ?
And all that based upon your incorrect assumption of what I was thinking. Jeepers.
The biggest, and sometimes the very last, mistakes are often made by those who assume.
Now I've gone through every post here.
As far as I am aware JLH was one of, if not the first to publish use of the 0.22 ohm series output resistor with a solid state amplifier, but I am not aware I stated that JLH used a 0.22 ohm to avoid line reflections. Did I say 'line' reflections ?
I know I have stated his use of this resistor, and I have stated its uses, but I do not remember ever writing that JLH actually used it for reasons I have personally exampled, though I could have because I am struggling these days, and none of this makes it any easier.
Where did I link JLH and line reflections please ?
Did you study my Posts#174+175 in Lumanauw's prior thread ?
Is there not evidence there of simulated loudspeaker system back-EMF induced amplifier distortion ?
Cheers .......... Graham.
I was not aware of the chip-amp recommendation.
Yes a thin cable can stop interface induced ringing, but sound better ? Only if you are very lucky !
When a thin wire is used, loudspeaker reproduction loses attack on transients and there is a tonal change.
If you use thin cable to a 2 or more way loudspeaker system that does not have near perfect impedance compensation it is likely to sound completely wrong, and not even the best graphic equaliser could compensate because the problem arises with waveform dynamics as well as amplitude.
I'd love to hear if that Creek sounds better with its own series 0.22 resistors.
Hi Jan,
Post#124. Hey I can't help what comes out when some question or a comment fires my neurones - we're all different.
I'm glad we can discuss this though, for I respect your views, but I am not capable of guessing what part of my understanding others think needs to be questioned !
Yes, that is my take too - a NFB loop controlled solid state amplifier is like a short circuit to the speaker.
I have already started going through this and saved it on my desktop, but the last couple of days leave me struggling for coherent thinking.
Hi Kanwar,
I cannot know or assume in advance what anyone else understands, especially when there are different ways of looking at fundamental theory.
So how can I be responsible for making someone understand, if they don't say what it is about my words, they doubt ?
Hi Jorge,
Re your first line question in Post#127. No ! Line reflections my pet theory ?
And all that based upon your incorrect assumption of what I was thinking. Jeepers.
The biggest, and sometimes the very last, mistakes are often made by those who assume.
Now I've gone through every post here.
As far as I am aware JLH was one of, if not the first to publish use of the 0.22 ohm series output resistor with a solid state amplifier, but I am not aware I stated that JLH used a 0.22 ohm to avoid line reflections. Did I say 'line' reflections ?
I know I have stated his use of this resistor, and I have stated its uses, but I do not remember ever writing that JLH actually used it for reasons I have personally exampled, though I could have because I am struggling these days, and none of this makes it any easier.
Where did I link JLH and line reflections please ?
Did you study my Posts#174+175 in Lumanauw's prior thread ?
Is there not evidence there of simulated loudspeaker system back-EMF induced amplifier distortion ?
Cheers .......... Graham.
- Status
- Not open for further replies.