Hi Mr.Mastertech,
You dont agree or you do agree, thats your way ...who cares.....ahh
By the way who told you that BJT's have higher damping factor than N-channel Vertical Mosfets......do you say under the influence of lateral mosfets scenario whose RDS is upto sometimes 5 times of vertical ones....
Cheers,
Kanwar
You dont agree or you do agree, thats your way ...who cares.....ahh
By the way who told you that BJT's have higher damping factor than N-channel Vertical Mosfets......do you say under the influence of lateral mosfets scenario whose RDS is upto sometimes 5 times of vertical ones....
Cheers,
Kanwar
Graham Maynard said:
Hi Graham,
The phase coherent crossover/loudspeaker damping for clean HF detail isn't the domain of BJT's alone it depends upon various designing parameters also ....The designer has to ascertain that such objectives are to achieved by the amplifier he has to design....The vertical mosfet also have Clean HF detail, but they rely on the driver parameter and designing along with the criteria of application also...
Cheers,
Kanwar
Hi Kanwar,
Of course design is important; and what suits bipolar does not automatically suit Mosfet, plus vice-versa.
Also optimising amplifiers for low forward distortion does not then automatically make them the best for coping with back-EMF, which can lead to higher reproduction distortion than is expected or might be acheivable with higher forward distortion circuitry.
Hi Mikeks,
I was rather tired when I wrote the comment you have highlighted.
Quite literally I had something to say that had been triggered by prior posting, but I could not get the words to come out right.
Looking back on them this morning I see that they do not say what I intended, so I am not surprised you question me.
Writing into a Forum can disadvantage the grammatical presentation of someone who has gone past his bedtime.
I wrote "low amounts of quadrature damping",
when I was actually wanting to write to Michael about;-
"low amounts of damping that is in quadrature".
Not quite the same, but at least clarified.
_______________________________________________
You also wrote;- "Please explain and/or define with exacting detail"
Mikeks. I am not a student sitting your two hour exam !
Yet you are obliging me to respond, and I wonder what your intended relevence is.
This is a open Forum. Have you been following this string, or are you just dipping in to challenge me ?
In Post#224 you *published* that I am making statements at variance with all known facts; and yet you are still expecting me to respond to you. ???
If you see something wrong then why not contribute *positively*, as we all try to do.
Normally when an error is spotted it is outlined in the hope that the contributor might usefully address, and if necessary, correct it and possibly update his understanding. None of us are anyways near to perfect, no matter how diligent our efforts in attempting to be.
In Post#224 you vaguely asked me to demonstrate for my statements relating to back-EMF. I can neither guess, nor assume on your behalf; therefore I do not know what exactly it is you are expecting me to explain. I did respond though, in Post#248.
Hi Mauro,
How's it going? Any more news?
Cheers ........... Graham.
Of course design is important; and what suits bipolar does not automatically suit Mosfet, plus vice-versa.
Also optimising amplifiers for low forward distortion does not then automatically make them the best for coping with back-EMF, which can lead to higher reproduction distortion than is expected or might be acheivable with higher forward distortion circuitry.
Hi Mikeks,
I was rather tired when I wrote the comment you have highlighted.
Quite literally I had something to say that had been triggered by prior posting, but I could not get the words to come out right.
Looking back on them this morning I see that they do not say what I intended, so I am not surprised you question me.
Writing into a Forum can disadvantage the grammatical presentation of someone who has gone past his bedtime.
I wrote "low amounts of quadrature damping",
when I was actually wanting to write to Michael about;-
"low amounts of damping that is in quadrature".
Not quite the same, but at least clarified.
_______________________________________________
You also wrote;- "Please explain and/or define with exacting detail"
Mikeks. I am not a student sitting your two hour exam !
Yet you are obliging me to respond, and I wonder what your intended relevence is.
This is a open Forum. Have you been following this string, or are you just dipping in to challenge me ?
In Post#224 you *published* that I am making statements at variance with all known facts; and yet you are still expecting me to respond to you. ???
If you see something wrong then why not contribute *positively*, as we all try to do.
Normally when an error is spotted it is outlined in the hope that the contributor might usefully address, and if necessary, correct it and possibly update his understanding. None of us are anyways near to perfect, no matter how diligent our efforts in attempting to be.
In Post#224 you vaguely asked me to demonstrate for my statements relating to back-EMF. I can neither guess, nor assume on your behalf; therefore I do not know what exactly it is you are expecting me to explain. I did respond though, in Post#248.
Hi Mauro,
How's it going? Any more news?
Cheers ........... Graham.
Hi Graham, have had little time, and am "buried" by a series of FFT that have performed and that have to analyse with calm.
In particular, have tried to isolate the dynamicses of pilotage of reactive load and real with current driver and voltage driver. They see well the behaviours of the "dynamic" damping and the "indestructible" nature of the Back_EMF . I am looking for of "paste" this test to the exhibited theories in this THread, but me maids of the time not to draw conclusions " too " make a mistake. To purpose of discussion, you show you some results:
This is a typical Voltage FFT @3Vrms to my old load (Filter LC 1Khz +6 ohm serie + 220uH serie), amp=DSLI (D.Self L. I.):
In particular, have tried to isolate the dynamicses of pilotage of reactive load and real with current driver and voltage driver. They see well the behaviours of the "dynamic" damping and the "indestructible" nature of the Back_EMF . I am looking for of "paste" this test to the exhibited theories in this THread, but me maids of the time not to draw conclusions " too " make a mistake. To purpose of discussion, you show you some results:
This is a typical Voltage FFT @3Vrms to my old load (Filter LC 1Khz +6 ohm serie + 220uH serie), amp=DSLI (D.Self L. I.):
Attachments
Obviously with resistive load this THD corresponds to that of voltage THD. How to never the current has a THD of 1% ? To the experts the answer ( I I know but per hour he doesn't expose theories...). Hour show you the same load but current driver ( partially, width bridge current about 100ohm Zout): First the voltage FFT:
Attachments
mauropenasa said:
Good job Mauro.
Did you keep the resistive dummy load or an actual speaker in your tests? If it is the former, current distortion is solely attributable to the back injection transformer magnetics. It should be interesting to repeat the test only with a speaker, without the back injection circuitry.
I go for the voltage mode drive for reasons I will post later. Let others voice their oppinion.
Rodolfo
current drive and DF...
sorry guys, I would say exactly the opposite!!
The voice coil is an e.m. linear motor, and as such it's movement is ~ proportional to the current flowing into it (please notice the "~": I say "circa"... I know it well it's all but a truly linear relation... but this does not substantially change things here, just makes them worse).
If you apply a voltage source, of course you get a current through the impedance of the voice coil, and that current moves it.
If the voice coil impedance was constant (even better if it was just a constant resistance...
) then there would be no problem... but...
but, unfortunately, the impedance of the voice coil of a loudspeaker is all but constant!
Thus, if you drive a loudspeaker with a voltage source, you get MORE distorsion out of the loudspeaker than if you drive it with a current source!
This is not an idea of mine. It's a matter of fact, predicted by theory and verified experimentally. See e.g. this interesting article:
You can find it here:
http://www.essex.ac.uk/ese/research/audio_lab/malcolms_publications.html
BTW, "my own" idea (currently proven "only" by listening tests with amplifiers which have a somewhat similar output characteristic... 😉 ) about this subject is a bit different.
I believe that possibly the best way to drive a speaker in order to obtain the best possible sound is neither by voltage nor by current, but by "power".
That is, my guess (a bit more than that, indeed...) is that a real loudspeaker would sound best when driven with a source which has approximately the same output impedance as the (average) input impedance of the loudspeaker.
For what regards the famous damping factor, if you look around there are plenty of articles (dating back to the '40s and '50s and then some... up to nowadays) which argue about the importance of achieving a critical (or at least near-critical) damping of the loudspeaker(s).
That is, to get the best sound out of a loudspeaker, you should neither underdamp NOR overdamp it, but should try to achieve a critical damping.
P.S.: for those afraid about "flapping cones", I would say... try it for yourself!
Up to now I have personnaly tried with several different speakers in several different setups. Using an amp which has a Zout close to their nominal impedence, I've heard (not so terrible) problems due to "low DF" only in one case... in which a couple of robust two-way Bass-Reflex speakers (capable of going down to some ~ 40Hz) were operated right against the back wall!
Thus, I claim that if you really need so much electrical damping from the amplifier, either you have chosen a bad position for your speakers, or your speakers are completely broken and/or ill-designed... and you'd better change them!
Jorge said:
sorry guys, I would say exactly the opposite!!
The voice coil is an e.m. linear motor, and as such it's movement is ~ proportional to the current flowing into it (please notice the "~": I say "circa"... I know it well it's all but a truly linear relation... but this does not substantially change things here, just makes them worse).
If you apply a voltage source, of course you get a current through the impedance of the voice coil, and that current moves it.
If the voice coil impedance was constant (even better if it was just a constant resistance...
) then there would be no problem... but...but, unfortunately, the impedance of the voice coil of a loudspeaker is all but constant!
Thus, if you drive a loudspeaker with a voltage source, you get MORE distorsion out of the loudspeaker than if you drive it with a current source!
This is not an idea of mine. It's a matter of fact, predicted by theory and verified experimentally. See e.g. this interesting article:
You can find it here:
http://www.essex.ac.uk/ese/research/audio_lab/malcolms_publications.html
BTW, "my own" idea (currently proven "only" by listening tests with amplifiers which have a somewhat similar output characteristic... 😉 ) about this subject is a bit different.
I believe that possibly the best way to drive a speaker in order to obtain the best possible sound is neither by voltage nor by current, but by "power".
That is, my guess (a bit more than that, indeed...) is that a real loudspeaker would sound best when driven with a source which has approximately the same output impedance as the (average) input impedance of the loudspeaker.
For what regards the famous damping factor, if you look around there are plenty of articles (dating back to the '40s and '50s and then some... up to nowadays) which argue about the importance of achieving a critical (or at least near-critical) damping of the loudspeaker(s).
That is, to get the best sound out of a loudspeaker, you should neither underdamp NOR overdamp it, but should try to achieve a critical damping.
P.S.: for those afraid about "flapping cones", I would say... try it for yourself!
Up to now I have personnaly tried with several different speakers in several different setups. Using an amp which has a Zout close to their nominal impedence, I've heard (not so terrible) problems due to "low DF" only in one case... in which a couple of robust two-way Bass-Reflex speakers (capable of going down to some ~ 40Hz) were operated right against the back wall!
Thus, I claim that if you really need so much electrical damping from the amplifier, either you have chosen a bad position for your speakers, or your speakers are completely broken and/or ill-designed... and you'd better change them!
ingrast said:
and that was the problem... 😀
seriously, NFB (expecially overall feedback) and vacuum tubes does NOT work well togheter at all. You should try to listen to a good, well designed, completely feedback-less modern SET amplifier (say one built around an 845 DHT with ~ 1.2KV anode supply, some 10Kohm reflected anode load, an adequate -that is, *STRONG*- driver, ...) connected to a pair of loudspeakers with decent sensitivity and impedance figure...
...and I bet you'll never want to listen to solid state ever again! 😉
UnixMan said:
By that time, standard practice was single ended class A pentode no feedback. Push-pull class A or AB was for higher end, and NFB more or of a curiosity. Perhaps your remark has to do with the fact a coupling transformer is a nasty piece of electronics that severily limits the amount of practical feedback while keeping a decent phase margin.
As to a experience pure vacuum tube class A sound as you propose, I have had no chance. May be you are right but I see no technical basis to expect that.
Rodolfo
Re: current drive and DF...
Very interesting article. After a (very cursory) review, it seems current drive cannot be applied unless either a second order equalizer is prepended to the (flat) amplifier, or motion feedback is used.
The first method suffers (as the authors admit) from the same voice coil heating effects as voltage drive, while the second of course implies specially designed speakers and tricky stability issues.
Rodolfo
UnixMan said:
Very interesting article. After a (very cursory) review, it seems current drive cannot be applied unless either a second order equalizer is prepended to the (flat) amplifier, or motion feedback is used.
The first method suffers (as the authors admit) from the same voice coil heating effects as voltage drive, while the second of course implies specially designed speakers and tricky stability issues.
Rodolfo
Graham Maynard said:
Not my intention to post a 'challange' for its own sake...Graham...
Members are hoping to learn from shared insights such as yours...
Such 'insights' are of no value if they are not backed up by cohesive explanation, founded on proven first principals.....
You see, damping is a ratio....
Now...what exactly is it in 'quadrature' with?
Moreover, if you can demonstrate its existence, what is its relevance in the scheme of things....?
Cheers.
UnixMan
From strictly the voice coil aspect, you may be right.
But pls tell speaker manufacturers that they should start designing speakers for current drive, and such tricks as using rising voice coil impedance to compensate for rising freq response of the cones cannot be used anymore, etc.
From strictly the voice coil aspect, you may be right.
But pls tell speaker manufacturers that they should start designing speakers for current drive, and such tricks as using rising voice coil impedance to compensate for rising freq response of the cones cannot be used anymore, etc.
UnixMan said:
If it wasn't for vacuum tubes and overall FB - ever heard of Williamson? - we wouldn't be at this forum today.
It's interesting that before Williamson, amps used tubes and had no feedback, and nobody (or the vast majority of audio freaks) said the new amplifier sounded bad...
Re: Re: current drive and DF...
In the classic mass-controlled driver (your average voice coil
piece), the first order approximation of response is flat above
resonance up to the point where the diaphragm size is
comparable to the wavelength. In this region you find that
output is directly proportional to acceleration, and so is directly
proportional to the current through the voice coil. As a result
it should not need equalization. Motional feedback gives
damping and control at the lower frequencies, but is not very
helpful above there. I see it as optional for high sensitivity
drivers if you are willing to bend the concept a bit in order to
have some damping factor, say on the order of 1 or so.
ingrast said:
In the classic mass-controlled driver (your average voice coil
piece), the first order approximation of response is flat above
resonance up to the point where the diaphragm size is
comparable to the wavelength. In this region you find that
output is directly proportional to acceleration, and so is directly
proportional to the current through the voice coil. As a result
it should not need equalization. Motional feedback gives
damping and control at the lower frequencies, but is not very
helpful above there. I see it as optional for high sensitivity
drivers if you are willing to bend the concept a bit in order to
have some damping factor, say on the order of 1 or so.
Hi Rodolfo, I have performed this test without any external signal, and the source of THD that sees is produced from the magnetic perturbations of the choke that I have used, and increases with the resonance of the LC.
I have used this setup because the resonance ( maximum point of Back_EMF and delta phase) is to an easy frequency by real-time analyse (without soft/hard -filtres ).
To demonstrative title I public 1 FFT with a 12V Motor load (7ohm Z & more Back_EMF @ all frequencies) Noted the simile...
Hi UnixMan, your link ( and your contribution ) is a lot of useful, because me allows to deepen of the theories that I had already realized, but had not read...
I like that my impressions on the better way for driver of loudspeakers is shared by other.
Ciao
Mauro
FFT1: Motor 3Vrms 1Khz (not resonance point, but some reactive load...)
I have used this setup because the resonance ( maximum point of Back_EMF and delta phase) is to an easy frequency by real-time analyse (without soft/hard -filtres ).
To demonstrative title I public 1 FFT with a 12V Motor load (7ohm Z & more Back_EMF @ all frequencies) Noted the simile...
Hi UnixMan, your link ( and your contribution ) is a lot of useful, because me allows to deepen of the theories that I had already realized, but had not read...
I like that my impressions on the better way for driver of loudspeakers is shared by other.
Ciao
Mauro
FFT1: Motor 3Vrms 1Khz (not resonance point, but some reactive load...)
Attachments
Re: Re: Re: current drive and DF...
Admittedly I was not precise in that the observed problem was about the resonance zone, where equalization is applied.
With respect to the velocity or acceleration sound pressure dependence, I should like to think and read a little more.
It is true that for a flat piston, far field sound pressure is proportional to piston velocity times frequency, which for sinusoidal signals happens to be the derivative of velocity and thus acceleration, but then the radiation pattern is non isotropic and frequency dependent so I have doubts for the moment about which is the relevant control variable from the standpoint of accurate reproduction. Remeber that for far field, the divergence theorem relates pressure and velocity through density and elasticity (impedance).
Rodolfo
Nelson Pass said:
Admittedly I was not precise in that the observed problem was about the resonance zone, where equalization is applied.
With respect to the velocity or acceleration sound pressure dependence, I should like to think and read a little more.
It is true that for a flat piston, far field sound pressure is proportional to piston velocity times frequency, which for sinusoidal signals happens to be the derivative of velocity and thus acceleration, but then the radiation pattern is non isotropic and frequency dependent so I have doubts for the moment about which is the relevant control variable from the standpoint of accurate reproduction. Remeber that for far field, the divergence theorem relates pressure and velocity through density and elasticity (impedance).
Rodolfo
mauropenasa said:
I suspected that Mauro. May I ask you to graph voltage and current THD with a speaker and without the choke and back injection? Just to have a look at the relevance of speaker impedance nonlinearities.
Rodolfo
Were working at it was described theories by this articles (UnixMan link xJ12 & J14) , on the basis only of the observations of this THread and from any mathematical relationships tricks on the models of loudspeakers...
I have received variue critical for this, and invite the "critics" to read this articles (UnixMan link xJ12 & J14) to understand that were the matters that me convinced...
I believe that this means that this forum is a "real stimulus" to study the launch problems of a amp, and that pushed who's who by a honest will of analysis not doesn't need " hide " after the theories.
Graham diatribe:
I make a mistake me much often, but the gotten damping in a circuit activates is "sensitive" to the "dynamics" phase. Rodolfo has specified me that NFB does always her work, and is true, but the vectorial sums of U/I are not an opinion.
Normally a NFB (voltage) follow a rotation of phase " dosing " it output current to get the output voltage. A amp that works in current has not this behaviour. Not doesn't need it current increase because it limits to follow "dinamic" phase ( only it reference point).
If we combine the two things in a real amp, in the non linears zones the real damping it may improve or worsen in operation of a funtion of phase shift.
This is what I associate to the concept of " damping quadratures " of Graham. I hope of have understanded well... 🙄
Hi Rodolfo, I look for some FFT and you show them...
Ciao
Mauro
- Status
- Not open for further replies.