Vacu, i will publish measurements and pictures of my system on my own MPL thread in the next half hour. There you will see that my speakers are phase coherent or however you will call it. The Step response shows that very well. See also the excellent low energy storage and fast settling in the ETC. For excellent time domain behavior you need a flat response too. Off cause you can make a flat speaker with a lot of phase rotations. You simply see the all pass term. For an interesting article how the brain can suppress the room see the Linkwitz AES paper.
By the way, Zobel your speakers. It works in my system. Recommended to try at least.
By the way, Zobel your speakers. It works in my system. Recommended to try at least.
Hi Joachim,
From what I saw about your system thus far was very interesting. The full range you are using is quite something special if that FR is even approximately correct. One of the systems I would definitely would like to hear.
vac
From what I saw about your system thus far was very interesting. The full range you are using is quite something special if that FR is even approximately correct. One of the systems I would definitely would like to hear.
vac
John, as far as i know this is a solid core ribbon cable, so only one solid conductor per side. I will measure the DC impedance tomorrow but is is quite wide so should be low. It sounds at the moment so good and i have taken out the cable the last days so many times that i am afraid to destroy the magic moment.He laquerizes the cable and then isolates the back and forth with natural rubber gum as far as i know. The outer jacket is untreated cotton.
The whole schebang is rather light and flexible.
Here is the Linkwitz paper. It´s good if not totally complete. Linkwitz is not able ( over headphones ) to hear the phase distortion of his 4th order Q 0.5 crossover network although he told me in a privat tutorial that it "may" be audible in the bass. Manger told me other wise ( "it´s the transients in treble you know, you stupid...."). Ploblem is with head phone hearing and sitting in the sweet spot that transients in both channels localize dead in the middle so Daniela and i proposed an off axis listening test. Nobody has taken up that load so far.
http://www.linkwitzlab.com/AES'07/AES123-final2.pdf
The whole schebang is rather light and flexible.
Here is the Linkwitz paper. It´s good if not totally complete. Linkwitz is not able ( over headphones ) to hear the phase distortion of his 4th order Q 0.5 crossover network although he told me in a privat tutorial that it "may" be audible in the bass. Manger told me other wise ( "it´s the transients in treble you know, you stupid...."). Ploblem is with head phone hearing and sitting in the sweet spot that transients in both channels localize dead in the middle so Daniela and i proposed an off axis listening test. Nobody has taken up that load so far.
http://www.linkwitzlab.com/AES'07/AES123-final2.pdf
You are lucky: my wife is the one who brings money into the family, so I have to use vacuum cleaner... However, I am lucky too: I took my speaker cables from the old cleaner that we discarded.
Now I have to run a refrigerator through them for a while, for proper cryo-treatment... I can do that while my wife is at work!
Vacu, i use a mix of passive EQ and line filtering. That way i get the driver pretty flat.
Watch my MPL thread if you want to see more. I think this does not fit in here so well but i must say without good speakers and a good room spending a lot on cables is a waist of energy and resources.
Watch my MPL thread if you want to see more. I think this does not fit in here so well but i must say without good speakers and a good room spending a lot on cables is a waist of energy and resources.
In fact, i'm divorced. Because i was spending too much time in burning-in wires. Now, i am obliged to bring the money AND to use the vacuum cleaner. On my point of view, you are the Lucky one.
If your supply use a neutral and a phase, don't forget to invert the plug half of the time, in order your cable is equally burned-in for positive and negative signals. (I have read that on a "high end" "state of the art" audiophile cables site).
nb: I wonder why there is so few female involved in Hifi, despite some of them seems to love burning things (food or shirts).
Depends, I still contend that there are not enough free parameters for the two methods to diverge at all. That is given Zo, length, etc. the lumped analysis would be like the design of T or Pi network L/C filters where there exists the same concept of characteristic impedance. Furthermore I bet the buildup of energy storage due to reflections is exactly mimiced in the energy stored in the passives. The same understanding applies to both (at tiny fractional wavelengths at least).
For me including the mention of the multiple reflections just makes things confusing.
Problem is not many people have heard a coplanar, coaxial, phase linear speaker. The Quad ESL63 comes close but is purely executed ( lots of mechanical resonances and obstacles ) and has not much dynamic range or bass to talk of. The Manger Transducer itself has a lot of FRD and distortion problems but was the first speaker i heard back in the 70th that had this ability ( although only at low volume ) to totally disappear. Depth perspective was also sensational. I could hear things 30 meters away. Spooky. What worked also for me was the Beveridge, although heavy modified in Peter Forsells witch kitchen. The original had a really poor amplifier and only one special guy made the panels ( i met him, a professional photographer, but forgot the name ) with any kind of reliability and pair matching. Then came the Apogees. This are both phase linear line sources and work only in big rooms. What then can happen is astounding. Maggies have a glimps of it at minimal outlay and other full range electrostatics. Always the same effect, this reach out and touch it feeling. There may be no scientific evidence that passes an ABX test but for me the anecdotal evidence is too massive to ignore the fact that phase coherence is important for correct imaging. And so it goes as JC would say.....
Joachim,
This underscores my point that Amplifiers, Cables and Speakers should be seen as a system and designed in ways that make sure things work together.
We may design Amplifier and speaker to produce optimal results with "Zipcord like" cables (the like because the cables should be solid core), which have a well known impedance and lumped LCR behaviour (this means networks integrated into amplifiers and speakers and using tweeters with appreciable inductive impedance rise).
Or we may design a cable to work optimally at a wide range of source and load impedance (means integrating a range of networks into the cables).
Or we may design our amplifiers and speaker to give correct result with a specific cable (Naim like).
As I use speakers that have tweeters with very low inductive impedance rise, this takes care of correctly terminating my very low charateristic impedance cables. All my amplifiers have sufficient inductance and resistance between the feedback take-off point and the output to be stable with such high capacitance cables.
So essentially, items in my system are designed to work with my favourite cable... 😀
Ciao T
This underscores my point that Amplifiers, Cables and Speakers should be seen as a system and designed in ways that make sure things work together.
We may design Amplifier and speaker to produce optimal results with "Zipcord like" cables (the like because the cables should be solid core), which have a well known impedance and lumped LCR behaviour (this means networks integrated into amplifiers and speakers and using tweeters with appreciable inductive impedance rise).
Or we may design a cable to work optimally at a wide range of source and load impedance (means integrating a range of networks into the cables).
Or we may design our amplifiers and speaker to give correct result with a specific cable (Naim like).
As I use speakers that have tweeters with very low inductive impedance rise, this takes care of correctly terminating my very low charateristic impedance cables. All my amplifiers have sufficient inductance and resistance between the feedback take-off point and the output to be stable with such high capacitance cables.
So essentially, items in my system are designed to work with my favourite cable... 😀
Ciao T
What can i say: i'm deseperate.
Can-you explain what you mean by "phase linear" ? (As long we will use material with more density than air to create air pressure ?)
From the very beginning of audio reproduction, people attemped to find a "one way" tranducer. All the way to make movements from an electrical signal was tried. From electrostatic devices (Quad) to planar systems (Gego) without forgetting conventional loudspeakers like Supravox or Bose (electronic side) attempts . No one never worked properly for physical reasons, OOT here.
Just a nice link about wide band loudspeakers legends:
http://www.dcx2496.fr/docs/Le%20large%20bande.pdf
Yes, some people where able to build "agreable to listen" one way speakers. But hifi ? Far away.
Sorry to say that, but, in the actual situation of the technology, your words: "phase linear speaker" says the contrary that the previous words you use in the same sentence.
Coaxial where primary designed by Tannoy (Very bad studios monitors). It does not worked well. Because it is impossible to physically align the loudspeakers in phase (numeric filters with delays can help on that, now), and because the treble loudspeaker use the bass cone as a horn. This does not work for several reasons. One of them is: or the cone is designed to be rigid enough, and his form will not be accurate for a horn, or it will fractionate at a too low frequency.
If you put the treble traducer in front of the cone, the obstacle it makes will create a lot of disturbance.
Electrostatics devices ? We all knows here why it is impossible to produce a *consistant* loud reproduction this way (Dynamic enough) . Ribbons suffers the same problem for the same reasons.
Planars (Geco)? It was a nice attempt, not enough efficiency, very low impedance, who even remember that, todays ?
Other attempts (some are intelligent) like Ionovac, Heil,Apogee, who remember them todays ?
And, about enclosures, What can we says ? Open baffles introduce too much room resonances problems, closed enclosures loose too much efficiencies, stays both bass reflex and acoustic lines, both with their own advantages and inconveniences.
To make a long story short, if you want to design an enclosure with a good linearity, both in phase and in response curve, not forgetting efficiency and distortions, you are stuck in multi ways systems. 2 ways are very acrobatic to design, 3 is OK.
Less you can have ways and crossovers, better it is.
Nowadays, with the help of numeric filters, we can help a little with the phase and response curves problems, it can make some nice looking measurements curves, but, at the end, does not change a lot when you listen at.
If you want to get enough basses, with enough level, you are stuck with big diameters electro dynamic loudspeakers.
If you want to get medium and treble with high efficiency and low distortion, you are near to be stuck with motors and horns.
For all those high frequencies out of our listening range, you can play with all the technologies you want to satisfy your marketing manager with poetic assertions.
Looks like you are living in a world of mystery and fairy tails, using esoteric cables as snake oil cooled equalizers .
If it makes-you happy, why not ?
Last edited:
A link to those unknown interesting "Orthophase" loudspeakers:
L'Orthophase, histoire et volution (E.Spiteri)
( In this article is a reference to the Ionovac, the revolutionary tweeter of dr Klein i was talking about.)
The inventor, Gerad Gogny was a friend of mine, i had tried to help him, and the manufacture i was working in had bough his patent in seventies on my recommendations, but never those strange loudspeakers had any commercial success. Too much expensive to build, low efficiency, low impedance, too far the market's normalization.
Gerard Gogny was the first, too, to build a double moving coil loudspeaker,one as motor, one as a sensor. The sensor was used in the amplifier feedback loop for electronic increased control of the loudspeaker.
That could be done todays with some light inertia captors, with less side effects (moving coil inertia, inductive coupling, etc.)
Last edited:
A correct t-line model is exact. The lumped analysis gets better as the number of elements is increased, they indeed have to converge. In addition, the concept of physical propagation becomes more real as the number of elements increase, it is easy to visualize the energy travelling towards the speaker.
The real questions are: how many elements are required to get "close". In general, I would guess one set is normally sufficient, but of course depends on one's definition of close.
Bet?? You bet it is mimicked??? It absolutely IS mimicked..😀...
As I've said many times, using the t-line analysis provides insight into the system's behaviour. It provides an excellent feel as to what is going on.. A simple LC model does not intuitively show the zero delay associated with matching the line to load impedance the way the t-line does. Nobody looks at L and C and says...hey, those two numbers together cause the settling delay at the load to disappear leaving just the prop velocity...
The cable Z certainly does..
And, as another positive, the t-line understanding provides insight into the nature of the relation between L and C. This is how I determined that Juachim's numbers were incorrect. His LC product was consistent with a t-line that had a dielectric constant of 24. Certainly incorrect...
A very important thing to remember is that energy travelling down a wire pair or t-line at the line's prop velocity can only have the characteristic V = Z*I. To believe that the energy that gets to an 8 ohm load via a 100 ohm cable travels at prop velocity is just incorrect..it does not. And, the lumped element numbers do not "reflect" that idea.
To me, it does not. It provides a clarity into other aspects of the system behaviour.
Cheers, John
Last edited:
Measuring a circle in order to determine the Pi number ?
Your scientific approach seems to be the only good one as long we are interested in line propagations of hf signals, for sure.
When it comes to the real world, connecting an enclosure to an amplifier, both approachs are equal: Out of scale. As long we will have loudspeakers with +/-10db of curves accidents and hundred of degrees of phase accidents.
It is like working on the painting of a formula one in order to ameliorate the skin effect at hight speed air penetration.
Better to work on tires, suspensions, motors and aerodynamic forms, if you want to win the race. Don't you think ? And just worry about the good looking of your painting for your sponsor's satisfaction.
Hi,
Yes, I agree. I do not know where start in all this muddle of misconceptions, misunderstandings and just plain misinformation. Let us take them one by one.
One way indeed has limitations.
However, a so-called "one way" can cover around 50...100Hz to 10KHz without acoustical and mechanical crossover and hence with a real minimum phase behaviour. Once we equalise such a system flat within the desired bandwidth (using acoustic, electrical or electronic means) we will have a system that has no excess phase turn other than that produced by the systems acoustic rolloff.
By adding subwoofers and supertweeters with suitable filters (generally acoustic first order types, which places quite some demands on the drivers) we can extend the "one-way" to a wider bandwidth while not compromising the phase behaviour much.
This we agree upon. A correctly implemented "one way with help in bass and treble" system offers a immediacy, directness and "open window" feeling that makes most conventional so-called "HiFi Speakers" sound muddled, indistinct and sorely lacking in ANY fidelity.
In fact, with respect, I have for a long come to associate supposed "HiFi" System with a mechanical, artifical and distinctly un-lifelike sound, in essence the polar opposite of what the result of a "High Fidelity" system should be.
This is neither the place nor time to debate the underlying technical issues (non the least because the various technocrates here have vested interests in preventing discussions of alternate approaches).
Hmm, let us for a moment take a purely theoretical animal, a 30Hz-15KHz +/-3dB "one way speaker". What is the phase response of such a speaker? Shall we call a speaker "phase-linear" if it's excess phase compared to the above is below a given target point? Say 30 Degrees?
First, Tannoy was not the only or main manufacturer of Coaxial systems, instead Eckmiller (Germany, late 30's) and James Bullock Lansing (USA, 1943) have clear priority, Olson (for RCA) had his coaxial (LC-1) around the same time as Tannoy. As to did the Tannoys work well or not, that is a matter debate and model.
Where do we start. The original Tannoys (up to Monitor Red) had a design of crossover and mechanics between cone driver and compression driver that caused together exactly 1/2 of the wavelength of the crossover frequency delay for the tweeter, with respect to the woofer. Invert the tweeter and all falls into place.
Sadly, later designs of Tannoys (Monitor Gold and later) first altered the crossover (to a higher order one to improve power handling) and later the mechanical geometry when converting from the deep alnico magnets to shallow ferrite magnets (somewhere in the HPD line), which destroyed this excellent behaviour regarding phase- and impulse-response.
Later Tannoy and Urei with Altec Coaxials used electrical delay lines to correct the phase/impulse response, the East German RL-900 coaxial active 3-Way Monitor (the Granddaddy of the the ME Geithain Monitors) used a midrange/treble assembly suspended in front of a 40cm Woofer and electronic (analogue) delay circuits to correct the arrival times.
Both the east German 12" Coaxial Monitors of the later years and the West German Isophon Orchster coaxials used tweeters placed before the cone. Altec/Urei had large horns in front of the bass driver, in these cases as well as the MEG's often foam was used to control the acoustic behaviour.
As to "the horn does not work", just measure a Tannoy monitor red one day (crossover will need refurbishing, sometimes the drivers as well). Quite a few years back some people did some extensive (including cepstral) analysis of many different horns combined with listening tests.
Tannoy coaxials where considered the best in terms of audible performance and showed the cleanest measurements.
It is entirely possible... You would need to go DIY though.
Large enough ribbons are hard to make.
Not sure what you are on about? Perchance Apogee, the reason that Krell Amplifiers where "invented"?
Maggies for example have been magnetic planars for many decades and are very well regarded by many.
I do, more than that as well...
!!!??? Boy have you got your lions crossed...
Open Baffles do not "introduce too much room resonances problems", they actually do the opposite...
Yes such systems are next to impossible to make to offer even as good a phase response as the above mentioned "one way plus helper bass & treble", though may technically claim such a system "3-Way" (I tend to consider them 1.5.5 Ways).
Well, if we use a simple System that is coincidental (aka coaxial or one-way) we can get a lot of mileage out of digital equalisation and time alignment.
Ciao T
Yes, I agree. I do not know where start in all this muddle of misconceptions, misunderstandings and just plain misinformation. Let us take them one by one.
One way indeed has limitations.
However, a so-called "one way" can cover around 50...100Hz to 10KHz without acoustical and mechanical crossover and hence with a real minimum phase behaviour. Once we equalise such a system flat within the desired bandwidth (using acoustic, electrical or electronic means) we will have a system that has no excess phase turn other than that produced by the systems acoustic rolloff.
By adding subwoofers and supertweeters with suitable filters (generally acoustic first order types, which places quite some demands on the drivers) we can extend the "one-way" to a wider bandwidth while not compromising the phase behaviour much.
This we agree upon. A correctly implemented "one way with help in bass and treble" system offers a immediacy, directness and "open window" feeling that makes most conventional so-called "HiFi Speakers" sound muddled, indistinct and sorely lacking in ANY fidelity.
In fact, with respect, I have for a long come to associate supposed "HiFi" System with a mechanical, artifical and distinctly un-lifelike sound, in essence the polar opposite of what the result of a "High Fidelity" system should be.
This is neither the place nor time to debate the underlying technical issues (non the least because the various technocrates here have vested interests in preventing discussions of alternate approaches).
Hmm, let us for a moment take a purely theoretical animal, a 30Hz-15KHz +/-3dB "one way speaker". What is the phase response of such a speaker? Shall we call a speaker "phase-linear" if it's excess phase compared to the above is below a given target point? Say 30 Degrees?
First, Tannoy was not the only or main manufacturer of Coaxial systems, instead Eckmiller (Germany, late 30's) and James Bullock Lansing (USA, 1943) have clear priority, Olson (for RCA) had his coaxial (LC-1) around the same time as Tannoy. As to did the Tannoys work well or not, that is a matter debate and model.
Where do we start. The original Tannoys (up to Monitor Red) had a design of crossover and mechanics between cone driver and compression driver that caused together exactly 1/2 of the wavelength of the crossover frequency delay for the tweeter, with respect to the woofer. Invert the tweeter and all falls into place.
Sadly, later designs of Tannoys (Monitor Gold and later) first altered the crossover (to a higher order one to improve power handling) and later the mechanical geometry when converting from the deep alnico magnets to shallow ferrite magnets (somewhere in the HPD line), which destroyed this excellent behaviour regarding phase- and impulse-response.
Later Tannoy and Urei with Altec Coaxials used electrical delay lines to correct the phase/impulse response, the East German RL-900 coaxial active 3-Way Monitor (the Granddaddy of the the ME Geithain Monitors) used a midrange/treble assembly suspended in front of a 40cm Woofer and electronic (analogue) delay circuits to correct the arrival times.
Both the east German 12" Coaxial Monitors of the later years and the West German Isophon Orchster coaxials used tweeters placed before the cone. Altec/Urei had large horns in front of the bass driver, in these cases as well as the MEG's often foam was used to control the acoustic behaviour.
As to "the horn does not work", just measure a Tannoy monitor red one day (crossover will need refurbishing, sometimes the drivers as well). Quite a few years back some people did some extensive (including cepstral) analysis of many different horns combined with listening tests.
Tannoy coaxials where considered the best in terms of audible performance and showed the cleanest measurements.
It is entirely possible... You would need to go DIY though.
Large enough ribbons are hard to make.
Not sure what you are on about? Perchance Apogee, the reason that Krell Amplifiers where "invented"?
Maggies for example have been magnetic planars for many decades and are very well regarded by many.
I do, more than that as well...
!!!??? Boy have you got your lions crossed...
Open Baffles do not "introduce too much room resonances problems", they actually do the opposite...
Yes such systems are next to impossible to make to offer even as good a phase response as the above mentioned "one way plus helper bass & treble", though may technically claim such a system "3-Way" (I tend to consider them 1.5.5 Ways).
Well, if we use a simple System that is coincidental (aka coaxial or one-way) we can get a lot of mileage out of digital equalisation and time alignment.
Ciao T
Hi,
The Orthophase principle is copied from the mid 1920's Siemens & Halske / Telefunken "Blatthaller" invented by H. Riegger, used among other application for sound reinforcement at the 1936 Olympics.
Ciao T
The Orthophase principle is copied from the mid 1920's Siemens & Halske / Telefunken "Blatthaller" invented by H. Riegger, used among other application for sound reinforcement at the 1936 Olympics.
Ciao T
Loudspeakers are interesting, far more so than the wire you use to attach them to an amplifier. But.... the thread topic is "speaker cables," so let's stick to that as much as possible.
Hi,
P.G.A.H. Voigt was issued a British Patent in 1924 for such a motional feedback system using either a balanced bridge or sense coil and electronic amplification, thus pre-dating not only all other motional feedback schemes but also Harold Stephen Black's "invention" of negative feedback in 1927 which was later popularised by Bell/Wester Electrics.
Ciao T
P.G.A.H. Voigt was issued a British Patent in 1924 for such a motional feedback system using either a balanced bridge or sense coil and electronic amplification, thus pre-dating not only all other motional feedback schemes but also Harold Stephen Black's "invention" of negative feedback in 1927 which was later popularised by Bell/Wester Electrics.
Ciao T
My approach is applicable at all frequencies. And, it is essentially equal to the approximation which is lumped elements. They each have their advantages. Understanding both is the best for those who can do so.
How odd you should mention that. You need to watch "mythbusters". They dimpled the surface of an automobile to see if what occurs with golf balls is applicable to cars. It was, they enjoyed I believe, a 19% mileage advantage at 60 miles per hour, 100 kph.
Your assertion has been "busted"..😀
Cheers, jn
Phase linear is a loudspeaker that has the same acoustically measured phase response then the Hilbert transform of its amplitude response. You could also say that it is a minimum phase system. A non linear phase loudspeaker has an all pass term that comes mostly from the crossover. The raw drivers are mostly minimum phase to begin with except some pathologically challenged exotics. The complete speaker will of cause be bandwidth limited say from 20Hz to 20Kz with say a 12dB rolloff thereafter. That will cause phase rotation at the bandpass ends but an electronic amplifier with the same bandpass limit will cause the same phase rotation. So you could then say that a minimum phase speaker behaves the same then an amplifier with the same frequency response. A real speaker will of cause have some ripples in the micro structure. That is the reason i smooth acoustic responses with a 1/6 octave filter. The ear itself is approximate like that. To measure phase linearity i use the Step response function because it is more spread out in the time domain ( so better for visual inspection ) then the impuls response that covers much less real estate on the screen. You could also measure the group delay that should be flat in the range of interest. Many other types of measurement can be done ( Wavelets, Cosine Burst, CSD and so one ) but the physical reality is always the same so i stick to my Step response measurement.
See some measurements on the speakers i use. This is only the main panel that goes from 100Hz to 20Khz with a single wide band speaker. No need for a super tweeter in my case although i have a tweeter on the back side out of other reasons. I have active subwoofers though.
See some measurements on the speakers i use. This is only the main panel that goes from 100Hz to 20Khz with a single wide band speaker. No need for a super tweeter in my case although i have a tweeter on the back side out of other reasons. I have active subwoofers though.
- Status
- Not open for further replies.