SunRa said:
I sold my soul to the devil..............
Just kidding, its an older cone shape that is known to work good, this is the first time a seriously nice motor was ever put behind it that I am aware of. Usually 15" pro sound cones with 2" voice coils are used for the cheap junk. The cone material itself is pretty nice as well, just stiff enough and very well internally damped as you can see in the graphs. I went through quite a few formulations before picking this one and then quite a bit of experimenting with the doping of the cone body and surround. This design is almost 10 years old now actually............
Snake oil would have sold a hell of a lot better IMHO.
I had a total of 2 years driver experience when I designed the TD series and most of that was working part time at a speaker rebuild shop. I simply took the best of what I heard and combined it into something I wanted for myself. I have no formal training in this and do all my design drawings on napkins I will admit.
I had a total of 2 years driver experience when I designed the TD series and most of that was working part time at a speaker rebuild shop. I simply took the best of what I heard and combined it into something I wanted for myself. I have no formal training in this and do all my design drawings on napkins I will admit.
Very, very nice TD15M measurements - the Z measurement is especially noteworthy. Certainly not hard to cross over in the 800 Hz ~ 1 kHz region, where the large-format HF driver will have lots of headroom and very low distortion. The 97.8dB/meter sensitivity is especially welcome, since the midbass driver sets the overall system efficiency and headroom.
In the context of the university of Adelaide paper I don't suppose that it has occurred to Earl Geddes that the writer has not in fact read his book and might well have come to similar conclusions independently.
It should also be fairly obvious that the increasing slope of the horn causes the imaginary part of the field to extend further down the horn and the non parallel nature of the wavefronts causes scattering and thus evanescent waves locally, only if the local cut off frequency becomes the same as the evanescent wave frequency do they propagate.
rcw
It should also be fairly obvious that the increasing slope of the horn causes the imaginary part of the field to extend further down the horn and the non parallel nature of the wavefronts causes scattering and thus evanescent waves locally, only if the local cut off frequency becomes the same as the evanescent wave frequency do they propagate.
rcw
RCW's post has migrated over to the "Geddes on Waveguides" thread, a more suitable forum. Interesting that all of that massive optimization, which takes up a large part of the thesis, appears to result in waveguides strikingly similar to Dr. Geddes' OS waveguide. I hope the author of the thesis extends the technique to include visualization of time-domain response over a section of a sphere.
John & Nick, how goes the Alnico TD15M project? Something that is years away, or a bit nearer term - or maybe I should just keep quiet, and let you guys do your thing.
I must say the latest curves, combined with the efficiency figures, do look rather good, and represent a powerful argument for a simple system with a 15" midbass and a 18" long-excursion, low-distortion IB/OB medium-Q bass-fill driver.
The response of the TD15M is good enough the builder would have a reasonable choice between large and small-format compression drivers, with a wide range of possible crossover frequencies and potential horn/waveguide sizes. As Dr. Geddes mentions elsewhere, small-format compression drivers are the obvious choice if response to 16 kHz is desired. Not everyone is going to want a RAAL or Goto Unit supertweeter.
John & Nick, how goes the Alnico TD15M project? Something that is years away, or a bit nearer term - or maybe I should just keep quiet, and let you guys do your thing.
I must say the latest curves, combined with the efficiency figures, do look rather good, and represent a powerful argument for a simple system with a 15" midbass and a 18" long-excursion, low-distortion IB/OB medium-Q bass-fill driver.
The response of the TD15M is good enough the builder would have a reasonable choice between large and small-format compression drivers, with a wide range of possible crossover frequencies and potential horn/waveguide sizes. As Dr. Geddes mentions elsewhere, small-format compression drivers are the obvious choice if response to 16 kHz is desired. Not everyone is going to want a RAAL or Goto Unit supertweeter.
Lynn Olson said:
Lynn figure early 2009 to be realistic.
CLS said:Hi,
From the measurements of TD15M, I noticed the impedance peak at fs falls at about 28Hz, instead of 34.7Hz in the published data.
Does it mean it can play even lower?
I don't know if John used a softer spider for that test sample, its quite possible.
Thanks for the info, Nick!
Based on the previous comments, the TD15M has a higher-performance cone than the TD12M, and is the best all-around choice for the intended application of matching with a large-format compression driver with a crossover in the 700 Hz ~ 1 kHz region. Am I reading this correctly?
I'm also surmising that the TD15M will be the first Lambda driver to have a non-ceramic magnet - I hope the project turns out successfully!
Based on the previous comments, the TD15M has a higher-performance cone than the TD12M, and is the best all-around choice for the intended application of matching with a large-format compression driver with a crossover in the 700 Hz ~ 1 kHz region. Am I reading this correctly?
I'm also surmising that the TD15M will be the first Lambda driver to have a non-ceramic magnet - I hope the project turns out successfully!
Regarding the post above, I remember the discussion regarding the Everest Loudspeaker. Back there mr. McKinney told that he doesn't think that a 15" driver can go as high as 800-900Hz (and play well) and that is the main problem of that particular system.
I hope I am not remembering this the wrong way
I hope I am not remembering this the wrong way
Lynn Olson said:
Hi Lynn,
Yes, I'd say that the TD15M is going to be a slightly better choice than the TD12M. As you know, the 15" driver needs to move less than the 12" for any given SPL level, keeping it in a more linear operating range. It is slightly more efficient, meaning less power to get to a given level and less heat. The the 15" cone does seem to have more controlled and smoother breakup vs the 12" cone, but this isn't up until in the 3KHz or higher region anyway. I'll try to get some measurements on the 12" with the same conditions in the next day or two. In reality, if going up to 1KHz I think either option would be quite acceptable though.
As for the Alnico, the plan is to be able to match all parameters between the ceramic and alnico versions. This means designing the motors to be virtually identical in overall flux and flux distribution in the gap. All of the TD drivers use the same motor, so as we get the alnico vs ceramic settled, all TD drivers would all be available at the same time with the alnico option. I have a few different size alico rings I'm trying to get samples of here. Once they arrive I can get on things quickly, but unfortunately I can't do much about the speed in which the samples arrive.
John
SunRa said:Regarding the post above, I remember the discussion regarding the Everest Loudspeaker. Back there mr. McKinney told that he doesn't think that a 15" driver can go as high as 800-900Hz (and play well) and that is the main problem of that particular system.
I hope I am not remembering this the wrong way
It was Charles Hansen....and regarding drivers that use a rubber surround, the surround breaks up between 300-400 hz. Please don't take this as gospel as I am regurgitating this from another source.
Chris
Sorry for this off-topic, but I was re-reading the thread and because it seems we are discussing again and again recurrent topics (this is almost a hermeneutic circle ) , I thought it would be worthy to discuss it.
This post is taken from a couple of pages below, when talking about the best material for voice-coils and formers, in order to minimize power-compression:
Now, the post of Mr. Janowitz is interesting because contradicts this patent here: magnetic spider - inventor H.J. Luth . In fact this patent is behind the Hartley woofers and widerangers. A similar technology seems to be used on lowther drivers:Hi-Ferric . So is it possible that actually some iron uniformly distributed in the voice-coil helps it to better position it in the gap?
And according to the patent above, can it act as a spider?
Thank you!
) , I thought it would be worthy to discuss it. This post is taken from a couple of pages below, when talking about the best material for voice-coils and formers, in order to minimize power-compression:
Now, the post of Mr. Janowitz is interesting because contradicts this patent here: magnetic spider - inventor H.J. Luth . In fact this patent is behind the Hartley woofers and widerangers. A similar technology seems to be used on lowther drivers:Hi-Ferric . So is it possible that actually some iron uniformly distributed in the voice-coil helps it to better position it in the gap?
And according to the patent above, can it act as a spider?
Thank you!
Lynn Olson said:
They are actually the same pulp material, surround, etc.
This is an issue with EV and other midrange targeted pro sound drivers as well. The 15" option always has a better midrange, and the 10" is the worse. This is of course discounting the effects of beaming. It seems related to the diameters of everything, cone, coil, etc., and the depth of the cone. I know the coil diameter is a serious issue as the larger the coil diameter the worse the midrange gets all else being equal. Perhaps the 15" cone dimensions stack up to a perfect ratio of some sorts that is waiting to be learned. These cone shapes were developed long before I was born, I just juiced them up a bit for my design.
I find this approach a little troubling (adding a ferrous material to the voice-coil). The better loudspeaker designers have spent decades trying to reduce high-order distortion terms from the magnetic system, and from what I can see, shaping the fields in the gap (and around it) is not a trivial design exercise (John and Nick can attest to that). Surely adding (nonlinear) ferrous material to the voice coil makes analysis of the problem more difficult, not easier.
One area where loudspeaker designers frequently go astray is focussing too heavily on large-signal problems - the amount of 2nd harmonic at hundreds of watts input, while not paying enough attention to 5th and higher harmonics at inputs of 1 watt or less. True, they are harder to measure, but they represent real-world listening levels a greater percentage of the time (on a statistical basis, peaks occupy only a small percentage of the overall envelope).
In addition, higher-order harmonics are far more audible than low-order harmonics, particularly 2nd, which is nearly inaudible (in comparison to 3rd and higher-order harmonics). It was in the Fifties that D.E.L. Shorter of the BBC and Norman Crowhurst in the USA both suggested weighting harmonics by the square or even the cube of the order of the harmonic, in order to better represent their audibility. It wasn't done then because the computational problems, but is much easier now.
The whole issue of non-metallic (ceramic) versus metallic (AlNiCo, neodymium, and field-coils) magnets could very well come down to the proportion of high-order harmonics generated at moderate listening levels. In the absence of distortion measurements that go out to the 5th and 7th harmonic, though, this must remain speculative. It's certainly not a matter of speculation in amplifier design, where controlling high-order harmonics at low listening levels is of prime importance (Class AB distortion, thermal tracking of the bias system, removing non-linear switch contacts from inside the feedback loop, etc).
P.S. The comments about the diameter of the voice coil are relevant to the discussion here - since the voice coil in acoustical terms is a parasitic mass (it plays no direct role in radiating sound), the bigger it gets, the greater the opportunity for side-to-side rocking and other parasitic modes. Unfortunately, the smaller the voice coil, the hotter it gets from voice-coil heating, thanks to smaller area. The Lambda heat-sink, or the use of an underhung voice coil, makes a real difference, since the VC can radiate its heat into a large adjacent heat sink.
It may seem like quibbling over 1 dB differences in efficiency is a minor matter, but every 1 dB loss of efficiency represents another 26% increase in voice coil heating (for the same SPL at the listening position), as well as another 26% more power required from the amplifier (which mostly goes to heat the voice coil, not make sound).
One area where loudspeaker designers frequently go astray is focussing too heavily on large-signal problems - the amount of 2nd harmonic at hundreds of watts input, while not paying enough attention to 5th and higher harmonics at inputs of 1 watt or less. True, they are harder to measure, but they represent real-world listening levels a greater percentage of the time (on a statistical basis, peaks occupy only a small percentage of the overall envelope).
In addition, higher-order harmonics are far more audible than low-order harmonics, particularly 2nd, which is nearly inaudible (in comparison to 3rd and higher-order harmonics). It was in the Fifties that D.E.L. Shorter of the BBC and Norman Crowhurst in the USA both suggested weighting harmonics by the square or even the cube of the order of the harmonic, in order to better represent their audibility. It wasn't done then because the computational problems, but is much easier now.
The whole issue of non-metallic (ceramic) versus metallic (AlNiCo, neodymium, and field-coils) magnets could very well come down to the proportion of high-order harmonics generated at moderate listening levels. In the absence of distortion measurements that go out to the 5th and 7th harmonic, though, this must remain speculative. It's certainly not a matter of speculation in amplifier design, where controlling high-order harmonics at low listening levels is of prime importance (Class AB distortion, thermal tracking of the bias system, removing non-linear switch contacts from inside the feedback loop, etc).
P.S. The comments about the diameter of the voice coil are relevant to the discussion here - since the voice coil in acoustical terms is a parasitic mass (it plays no direct role in radiating sound), the bigger it gets, the greater the opportunity for side-to-side rocking and other parasitic modes. Unfortunately, the smaller the voice coil, the hotter it gets from voice-coil heating, thanks to smaller area. The Lambda heat-sink, or the use of an underhung voice coil, makes a real difference, since the VC can radiate its heat into a large adjacent heat sink.
It may seem like quibbling over 1 dB differences in efficiency is a minor matter, but every 1 dB loss of efficiency represents another 26% increase in voice coil heating (for the same SPL at the listening position), as well as another 26% more power required from the amplifier (which mostly goes to heat the voice coil, not make sound).
chrismercurio said:
Yes rubber surrounds cause some amazing break ups. When I started down the low inductance road the first problem I found was that whatever was resonating in the driver showed mountain peaks in the impedance curve. I have never seen a worse set of impedance bumps than a prototype TD driver with a rubber surround. It took a special foam and special doping of the foam to get the impedance curve of our drivers smooth enough for my tastes, the foam itself had troubles also.
SunRa said:Sorry for this off-topic, but I was re-reading the thread and because it seems we are discussing again and again recurrent topics (this is almost a hermeneutic circle
This post is taken from a couple of pages below, when talking about the best material for voice-coils and formers, in order to minimize power-compression:
Now, the post of Mr. Janowitz is interesting because contradicts this patent here: magnetic spider - inventor H.J. Luth . In fact this patent is behind the Hartley woofers and widerangers. A similar technology seems to be used on lowther drivers:Hi-Ferric . So is it possible that actually some iron uniformly distributed in the voice-coil helps it to better position it in the gap?
And according to the patent above, can it act as a spider?
Thank you!
I want nothing in the gap other than voice coil and air for any driver with an excursion of more than about 3mm peak to peak. I have rebuilt a few drivers with liquids and such in the gap and at higher excursion whatever junk they put in doesn't seems to stay and ends up in the spider and basically making a mess of everything.