Magnetar said:
It's not an open baffle, but it has been done (and yeah, it's not a ribbon either, but an AMT does have similar dispersion characteristics). This isn't my setup, but I used a Lambda TD12 with an AMT in a similar configuration and they sounded much better than they should have.
An externally hosted image should be here but it was not working when we last tested it.
Variac said:
Good point. Lynn is never less than a gentleman and to pollute his thread with hostility is not acceptable. As I have mentioned in the past, different threads have different attitudes, and this one requires civility.
LOL - I will not contribute to this thread anymore seeing it is too civil for legitimate argument for a simple thing such this driver mis-match 'recommendation' from a manufacture sales rep- The tread seems to be nothing but a long drawn out fantasy anyway. I don't think olson is going to get rich with it it either.
Why can't I upload files with my posts? I have emailed other moderators and they can't answer me. I'd like to start a thread with real measurements posted but you won't let me upload. Why is that?
IOW I can't attach files.
I have no problem with the tone of the discussion so far - it's been most interesting. I'm still doing research on the historical development of compression drivers since the advent of the Altec 288 in 1945, and the various twists and turns in the story after the late Seventies.
This was the transitional period when Altec, Emilar, and JBL developed several different solutions (Mylar surrounds, Titanium diaphragms, etc.) to address the problem of greatly higher power demands for PA and theatre use. The primary focus of the various solutions was durability, and many of them actually degraded sound quality. The hard part is disentangling marketing claims that were patently untrue from the genuine sonic advances.
Rather than "fix it after the fact", I like to know as much as possible about the underlying technology and its strengths and weaknesses. That's how I designed the Amity and Karna amplifiers - much research on triodes and their history, and what they were good at and what they weren't so good at.
All devices have favored regions of operation, and you have to know the design tradeoffs to find out what these were - it certainly won't be disclosed in marketing literature, which always carefully steers around the inherent weak points of the design. The best way to dig this out is research the original papers and patents, read between the lines for hidden assumptions and overgeneralizations, and talk to the designers, or people who knew them.
The majority of the real engineers I've met are all too willing to tell me about "what they really wanted do" before management shut down the project and rushed it into production. If they won't tell me their "war stories", that too tells me about the marketing vs engineering priorities of the vendor. If it's all positive and product is perfect - and I hear this from an engineer - then I run, don't walk, in the other direction.
The "compromised" areas of a given design can be very difficult to find with a "black-box" measurement on the completed, after-the-fact product. When you read the original papers, or speak to the designers, they point out the weak areas, and what to look for when measuring the finished product. This is the approach I prefer to take - get a grasp of the underlying technology, find where the weak points are, and steer the design away from those weak points. (Triodes like flat load-lines, transformer bandwidth is set by source and load impedances, sharp boundaries cause diffraction and reflection, etc. etc.)
Moving on to potential crossover issues, what controls the ratio of room energy to direct-arrival power is the directivity into a sphere, not merely the horizontal or vertical dispersion figures measured across a frontal arc. The large RAAL tweeter has close to 180-degree horizontal directivity, but the vertical isn't much more than 20~40 degrees as far as I can tell. It's the total power that's radiated into a sphere that matters, since the room reflections are the summation of emission in all directions (including the rear), not just a fan stretching across the listening area.
Rather than try a heroic 2-way between the 15-incher and the big RAAL, I'd add a large-format compression driver in the middle, covering, say 700 Hz to 7 kHz, which is well within what a 1.4" or 2" compression can do with much room to spare, and it also takes a huge load off the 15-incher and the ribbon tweeter. That way all of the drivers have an octave or more of spare bandwidth on each end, which can only improve the sound and relax the crossover design.
This was the transitional period when Altec, Emilar, and JBL developed several different solutions (Mylar surrounds, Titanium diaphragms, etc.) to address the problem of greatly higher power demands for PA and theatre use. The primary focus of the various solutions was durability, and many of them actually degraded sound quality. The hard part is disentangling marketing claims that were patently untrue from the genuine sonic advances.
Rather than "fix it after the fact", I like to know as much as possible about the underlying technology and its strengths and weaknesses. That's how I designed the Amity and Karna amplifiers - much research on triodes and their history, and what they were good at and what they weren't so good at.
All devices have favored regions of operation, and you have to know the design tradeoffs to find out what these were - it certainly won't be disclosed in marketing literature, which always carefully steers around the inherent weak points of the design. The best way to dig this out is research the original papers and patents, read between the lines for hidden assumptions and overgeneralizations, and talk to the designers, or people who knew them.
The majority of the real engineers I've met are all too willing to tell me about "what they really wanted do" before management shut down the project and rushed it into production. If they won't tell me their "war stories", that too tells me about the marketing vs engineering priorities of the vendor. If it's all positive and product is perfect - and I hear this from an engineer - then I run, don't walk, in the other direction.
The "compromised" areas of a given design can be very difficult to find with a "black-box" measurement on the completed, after-the-fact product. When you read the original papers, or speak to the designers, they point out the weak areas, and what to look for when measuring the finished product. This is the approach I prefer to take - get a grasp of the underlying technology, find where the weak points are, and steer the design away from those weak points. (Triodes like flat load-lines, transformer bandwidth is set by source and load impedances, sharp boundaries cause diffraction and reflection, etc. etc.)
Moving on to potential crossover issues, what controls the ratio of room energy to direct-arrival power is the directivity into a sphere, not merely the horizontal or vertical dispersion figures measured across a frontal arc. The large RAAL tweeter has close to 180-degree horizontal directivity, but the vertical isn't much more than 20~40 degrees as far as I can tell. It's the total power that's radiated into a sphere that matters, since the room reflections are the summation of emission in all directions (including the rear), not just a fan stretching across the listening area.
Rather than try a heroic 2-way between the 15-incher and the big RAAL, I'd add a large-format compression driver in the middle, covering, say 700 Hz to 7 kHz, which is well within what a 1.4" or 2" compression can do with much room to spare, and it also takes a huge load off the 15-incher and the ribbon tweeter. That way all of the drivers have an octave or more of spare bandwidth on each end, which can only improve the sound and relax the crossover design.
I apologize if I keep disappointing people who are expecting a construction project. I prefer to explore unusual, non-market-acceptable aspects of the audio world, turning up little scraps of historical research as I go along, trying out different concepts, measuring, listening, etc. Yes, I know I work very slowly. Everyone tells me that.
The "known" realm of commercial products is a lot less interesting, most of which simply copy each other, while spending the majority of R&D money on marketing - or more unhappily for the rest of us, aggressive legal departments.
Follow this link for an interesting story on a feisty vendor who fought back against this kind of intimidation. I particularly enjoy the detail that Monster Cable is transferring intellectual-property income out of the USA to a holding company in Bermuda - who would have guessed?
The "known" realm of commercial products is a lot less interesting, most of which simply copy each other, while spending the majority of R&D money on marketing - or more unhappily for the rest of us, aggressive legal departments.
Follow this link for an interesting story on a feisty vendor who fought back against this kind of intimidation. I particularly enjoy the detail that Monster Cable is transferring intellectual-property income out of the USA to a holding company in Bermuda - who would have guessed?
Construction project? Dammit, we want a kit. All options included; our choice of the latest-n-greatest wire and connectors; full user manual and documentation, 12 months warranty and a freephone help line staffed 24hrs a day. Oh, and high WAF as well. Now, what's the problem with that?
Keep up the good work!
Mike
Keep up the good work!
Mike
Lynn Olson said:
The info so far has been very interesting. Please keep it coming, Lynn, this is something that I am very interested in too
. References to articles, patents etc are also very much appreciated. And if there are specific articles you are looking for, just send me a PM and perhaps I can help.
Best regards,
Bjørn
So far so good. A few questions:
What is the normalized efficiency of the TD15M?
Is the cone profile similar to the Altec 416 or 515?
Which magnets are available - ceramic, Alnico, or Neodymium?
Is the voice coil underhung or overhung?
Price?
The provided curves for the TD15M/BMS MTM system show 8~12 dB peak-to-dip ripples in the 700 Hz to 2 kHz region, although very smooth above that - are these real or some kind of measurement artifact?
Attachments
Very Brief History of Compression Drivers
Bell Labs & Wente 1933 Compression Driver Pat# 2,037,187
Blackburn 1937 Compression Driver Pat# 2,183,528
Lansing & Hilliard Loudspeaker System for Theaters
Presented May 14, 1945 at the Technical Conference in Hollywood by Altec Lansing, describing the A-2 Theater Loudspeaker, and the 288 and 515 drivers.
Altec & Henricksen 1976 Tangerine Pat# 4,050,541
JK Hilliard on optimum diaphragm sizes
Jerry Hubbard of Altec on Titanium vs Aluminum
Jonas Renkus, co-worker of JK Hilliard at Altec, and founder of Emilar
Renkus, Emilar, and Radian (Part 2)
Don McRitchie Interview of Doug Button of JBL. This is worthy of very close reading. Some excerpts and commentary:
Not surprisingly, the large format drivers were intended for high output. However, there was a tradeoff in high frequency extension. Due to the higher mass of its large diaphragm, its response was limited to 10khz. The small format drivers, with their lighter diaphragms could extend response up to 15khz, but could not match the output levels of their larger counterparts.
The first significant innovation to these designs occurred in 1979 with JBL’s introduction of the diamond surround. These surrounds take advantage of a phenomenon called “parasitic resonance”. The aluminum surrounds have a natural resonant frequency. Forming the surround into a diamond pattern pushes the second resonance out in frequency to result in a broad band response. This allowed the large format drivers to match the frequency extension of the small drivers.
Note what this implies. The upper-HF response of large-format compression drivers is actually a broad region of resonance coming from the surround - and it is now standard practice in theaters to boost this region of declining response and increasing distortion, sometimes as much as 10 dB!
However, it was soon discovered that the diamond surrounds limited power handling. The diamond peaks are more susceptible to stress concentration and failure. This led JBL to pioneer the use of titanium as a diaphragm material in 1982. Titanium is an order of magnitude more resistant to fatigue failures than aluminum. With this new diaphragm, JBL now had compression drivers with exceptional output and extension.
However, it was recognized that there were compromises with the new diaphragms. Titanium does not have the internal damping of aluminum and thus has marginally higher distortion levels. The diamond surrounds, while extending frequency response, do so at the expense of transient response. Further, due to its lower stiffness, titanium goes into breakup at a lower frequency.
So what titanium has going for it is fatigue resistance and ability to withstand abuse, not anything to do with audio quality - in fact, it is worse, with more distortion, and goes into breakup at a lower frequency (4 kHz instead of 7 kHz). Combined with heavy equalization that is standard practice in theaters, now you know why the sound is so harsh. (Very few theaters are using beryllium diaphragms - Ti has become the industry standard, thanks to JBL's dominance of the theater market.)
In 1999, Doug Button began development on a new series of compression drivers that would result in the 435Be. With the 435Be, Doug wanted to address both the bandwidth and output requirements without compromise in distortion. The goal was to have a driver that was pistonic throughout its bandwidth and have extension that did not rely on parasitic resonance. The solution was in a different diaphragm material – beryllium. The use of beryllium in compression drivers was not new. The TAD division of Pioneer had been producing such drivers for many years. However, the approach and design objectives set for the 435Be were unique.
To ensure pistonic response, Doug specified a smaller 3" diameter. A beryllium diaphragm of this size ensured that breakup modes would be above 15.5khz. A thin layer of Aquaplas was applied to the back of the diaphragm to damp spurious resonances. The diaphragm was also light enough for extension to that frequency without parasitic resonance. Total moving mass was only 1 gram, with the diaphragm less than 0.5 grams. This was less than a third the weight of previous materials.
So, after nearly a half-century of using the 4" diaphragm in the JBL 375, they revert to a 3" diaphragm for their flagship product, just slightly larger than the 1945 Altec 288 size of 2.88".
Another unique feature of this driver is that it does not have a traditional throat. The phase plug terminates at the driver exit. It results in an effective flare rate of 550hz.
Previously, virtually every compression driver made had a 180hz flare rate whose origin dates back to the original AT&T Labs designs from the 1930's. This low rate was necessary to accommodate the low cross-over points used in early two-way loudspeakers. However, this low rate compromised high frequency performance. Given that there was no need for such low frequency output for the 435Be, the flare rate could be optimized to result in a 6db drop in second harmonic distortion.
The "throatless" design of the JBL 435Be has now been extended to many of their other professional drivers. Also interesting that traditional compression drivers have a short inbuilt flare rate of 180 Hz, with the choice of flare rate being little more than a historical artifact.
Bell Labs & Wente 1933 Compression Driver Pat# 2,037,187
Blackburn 1937 Compression Driver Pat# 2,183,528
Lansing & Hilliard Loudspeaker System for Theaters
Presented May 14, 1945 at the Technical Conference in Hollywood by Altec Lansing, describing the A-2 Theater Loudspeaker, and the 288 and 515 drivers.
Altec & Henricksen 1976 Tangerine Pat# 4,050,541
JK Hilliard on optimum diaphragm sizes
Jerry Hubbard of Altec on Titanium vs Aluminum
Jonas Renkus, co-worker of JK Hilliard at Altec, and founder of Emilar
Renkus, Emilar, and Radian (Part 2)
Don McRitchie Interview of Doug Button of JBL. This is worthy of very close reading. Some excerpts and commentary:
Not surprisingly, the large format drivers were intended for high output. However, there was a tradeoff in high frequency extension. Due to the higher mass of its large diaphragm, its response was limited to 10khz. The small format drivers, with their lighter diaphragms could extend response up to 15khz, but could not match the output levels of their larger counterparts.
The first significant innovation to these designs occurred in 1979 with JBL’s introduction of the diamond surround. These surrounds take advantage of a phenomenon called “parasitic resonance”. The aluminum surrounds have a natural resonant frequency. Forming the surround into a diamond pattern pushes the second resonance out in frequency to result in a broad band response. This allowed the large format drivers to match the frequency extension of the small drivers.
Note what this implies. The upper-HF response of large-format compression drivers is actually a broad region of resonance coming from the surround - and it is now standard practice in theaters to boost this region of declining response and increasing distortion, sometimes as much as 10 dB!
However, it was soon discovered that the diamond surrounds limited power handling. The diamond peaks are more susceptible to stress concentration and failure. This led JBL to pioneer the use of titanium as a diaphragm material in 1982. Titanium is an order of magnitude more resistant to fatigue failures than aluminum. With this new diaphragm, JBL now had compression drivers with exceptional output and extension.
However, it was recognized that there were compromises with the new diaphragms. Titanium does not have the internal damping of aluminum and thus has marginally higher distortion levels. The diamond surrounds, while extending frequency response, do so at the expense of transient response. Further, due to its lower stiffness, titanium goes into breakup at a lower frequency.
So what titanium has going for it is fatigue resistance and ability to withstand abuse, not anything to do with audio quality - in fact, it is worse, with more distortion, and goes into breakup at a lower frequency (4 kHz instead of 7 kHz). Combined with heavy equalization that is standard practice in theaters, now you know why the sound is so harsh. (Very few theaters are using beryllium diaphragms - Ti has become the industry standard, thanks to JBL's dominance of the theater market.)
In 1999, Doug Button began development on a new series of compression drivers that would result in the 435Be. With the 435Be, Doug wanted to address both the bandwidth and output requirements without compromise in distortion. The goal was to have a driver that was pistonic throughout its bandwidth and have extension that did not rely on parasitic resonance. The solution was in a different diaphragm material – beryllium. The use of beryllium in compression drivers was not new. The TAD division of Pioneer had been producing such drivers for many years. However, the approach and design objectives set for the 435Be were unique.
To ensure pistonic response, Doug specified a smaller 3" diameter. A beryllium diaphragm of this size ensured that breakup modes would be above 15.5khz. A thin layer of Aquaplas was applied to the back of the diaphragm to damp spurious resonances. The diaphragm was also light enough for extension to that frequency without parasitic resonance. Total moving mass was only 1 gram, with the diaphragm less than 0.5 grams. This was less than a third the weight of previous materials.
So, after nearly a half-century of using the 4" diaphragm in the JBL 375, they revert to a 3" diaphragm for their flagship product, just slightly larger than the 1945 Altec 288 size of 2.88".
Another unique feature of this driver is that it does not have a traditional throat. The phase plug terminates at the driver exit. It results in an effective flare rate of 550hz.
Previously, virtually every compression driver made had a 180hz flare rate whose origin dates back to the original AT&T Labs designs from the 1930's. This low rate was necessary to accommodate the low cross-over points used in early two-way loudspeakers. However, this low rate compromised high frequency performance. Given that there was no need for such low frequency output for the 435Be, the flare rate could be optimized to result in a 6db drop in second harmonic distortion.
The "throatless" design of the JBL 435Be has now been extended to many of their other professional drivers. Also interesting that traditional compression drivers have a short inbuilt flare rate of 180 Hz, with the choice of flare rate being little more than a historical artifact.
To summarize the rather confusing history of the late-Seventies and later period, all hell broke loose in the SR/rock concert and theater markets. The industry standard, the Altec 288 dating back to 1945 and used in theaters and PA systems worldwide, were failing under the onslaught of banks of multi-hundred-watt transistor power amps and massive equalization. (Remember, the 288 was designed in the days of 8 kHz optical sound tracks and 60-watt tube amplifiers with soft clipping characteristics.)
Altec engineering was being disrupted by repeated cycles of mindless corporate mergers that resulted in the eventual destruction of the company, a fate that JBL narrowly avoided. With limited resources, Altec first tried the Pascalite aluminum-alloy diaphragm, then the Mylar surround of the Symbiotik drivers (which replaced the 1945-era tangential surround of the 288).
The Symbiotik was developed further by Jonas Renkus as the new Emilar, and we see the continuation of the aluminum-diaphragm combined with the Mylar surround in the Radian product line. These products have found applications as an OEM supplier for the studio-monitor market.
JBL captured the rock and SR markets in the Seventies, and went on to replace Altec in the theater market in the Eighties. After Star Wars put Lucasfilms, Dolby Stereo, and THX certification on the must-have list for theaters, higher power levels and heavier equalization became the industry-standard. Titanium became the diaphragm material of choice for SR and theaters, thanks to high fatigue resistance, although the Emilar/Radian solution worked about as well, and didn't require titanium.
Beryllium was pioneered by TAD, but the TAD technology relies on extremely expensive (and slow) vapor-deposition technology. From what I understand, JBL buys beryllium as sheet foils from Brush-Wellman in the USA, sidestepping the expense and limited production quantities of the TAD technology. It's still not cheap, though - the 435Be (consumer, treated with Aquaplas damping) and 2435 (professional) compression drivers are well over a thousand dollars each. Don't expect any discounts unless you buy used.
The discussion of the merits of 4" vs 3" diaphragms, tangential vs diamond vs Mylar/plastic surrounds, circumferential vs tangential vs pepper-pot vs new-tech phase-plugs, Alnico vs ceramic vs field-coil magnets, Altec vs JBL vs TAD vs Emilar vs Vitavox S2's can go on endlessly. (I'm not even going to mention horns and waveguides, I've learned my lesson on that one.) Everyone in the high-efficiency biz has their (strong) favorites, and decades of marketing has created very strong brand loyalty. The underlying technology has gotten lost in a blizzard of buzzwords and marketing slogans.
It has to be remembered that SR, movie theater, and studio monitoring requirements are NOT the same. SR is all about power, precise coverage, and reliability. Sound quality comes well after the first three requirements. The movie theater business is dominated by a combination of cost control and buzzword compliance (THX, Dolby's latest, SDDS, etc.). Lip service is paid to sound quality but most theater operators are more concerned about the profitability of the snack-food concession.
Studio monitoring is about repeatability and "transportability" - the ability to make a mix that plays well on a wide variety of playback systems. Over time, the popularity of nearfield monitors that are console mounted, as well the proliferation of smaller "project" studios, has led to less use of traditional large soffit-mounted monitors. The lowly Yamaha NS10 is used more widely than you might think.
Altec engineering was being disrupted by repeated cycles of mindless corporate mergers that resulted in the eventual destruction of the company, a fate that JBL narrowly avoided. With limited resources, Altec first tried the Pascalite aluminum-alloy diaphragm, then the Mylar surround of the Symbiotik drivers (which replaced the 1945-era tangential surround of the 288).
The Symbiotik was developed further by Jonas Renkus as the new Emilar, and we see the continuation of the aluminum-diaphragm combined with the Mylar surround in the Radian product line. These products have found applications as an OEM supplier for the studio-monitor market.
JBL captured the rock and SR markets in the Seventies, and went on to replace Altec in the theater market in the Eighties. After Star Wars put Lucasfilms, Dolby Stereo, and THX certification on the must-have list for theaters, higher power levels and heavier equalization became the industry-standard. Titanium became the diaphragm material of choice for SR and theaters, thanks to high fatigue resistance, although the Emilar/Radian solution worked about as well, and didn't require titanium.
Beryllium was pioneered by TAD, but the TAD technology relies on extremely expensive (and slow) vapor-deposition technology. From what I understand, JBL buys beryllium as sheet foils from Brush-Wellman in the USA, sidestepping the expense and limited production quantities of the TAD technology. It's still not cheap, though - the 435Be (consumer, treated with Aquaplas damping) and 2435 (professional) compression drivers are well over a thousand dollars each. Don't expect any discounts unless you buy used.
The discussion of the merits of 4" vs 3" diaphragms, tangential vs diamond vs Mylar/plastic surrounds, circumferential vs tangential vs pepper-pot vs new-tech phase-plugs, Alnico vs ceramic vs field-coil magnets, Altec vs JBL vs TAD vs Emilar vs Vitavox S2's can go on endlessly. (I'm not even going to mention horns and waveguides, I've learned my lesson on that one.) Everyone in the high-efficiency biz has their (strong) favorites, and decades of marketing has created very strong brand loyalty. The underlying technology has gotten lost in a blizzard of buzzwords and marketing slogans.
It has to be remembered that SR, movie theater, and studio monitoring requirements are NOT the same. SR is all about power, precise coverage, and reliability. Sound quality comes well after the first three requirements. The movie theater business is dominated by a combination of cost control and buzzword compliance (THX, Dolby's latest, SDDS, etc.). Lip service is paid to sound quality but most theater operators are more concerned about the profitability of the snack-food concession.
Studio monitoring is about repeatability and "transportability" - the ability to make a mix that plays well on a wide variety of playback systems. Over time, the popularity of nearfield monitors that are console mounted, as well the proliferation of smaller "project" studios, has led to less use of traditional large soffit-mounted monitors. The lowly Yamaha NS10 is used more widely than you might think.
I think you can now see more clearly why I've been talking about supertweeters that come in around 7 kHz.
Compression drivers were originally designed for movie theaters with 8 kHz bandwidth soundtracks, a minimum of equalization, and moderate-sized vacuum-tube power amplifiers with what we now call soft clipping (the 20 dB of feedback of the Williamson amplifier came 2 years after the 288 was designed). Hard clipping didn't really arrive until the Crown DC300 starting seeing professional use - and compression drivers started to fail in large numbers.
There is even a brief mention towards the end of the Lansing/Hilliard paper that the 515 and 288 drivers give their optimum performance with an amplifier output impedance equal to the driver impedance (unity coupling) - a detail that's been lost in the shuffle once high-feedback amplifiers became standard.
More than sixty years after the 288 was introduced, I find it significant that the most heroic efforts, using the most exotic diaphragms and phase plugs, are only one octave better than the 288. That says a lot about the HF limitations of compression driver technology, because there certainly has no been shortage of time, money, and talent thrown at it. If the best minds in the industry have been "working the problem" for that length of time, I would gently suggest that the HF limitations of compression drivers are not going away any time soon. The technology is trying to tell us something, if we will but listen.
This is why I was posting earlier about finding out the underlying limitations of the technology, and using the (inherent) limitations to guide the design process, instead of vainly hoping for some kind of breakthrough. Breakthroughs in audio technology are actually very rare, and almost always come with a new set of limitations of their own.
My rather simple-minded process is to look at each technology, and find what it is good at, and not so good at. Ribbons are superb at pulse response and ultra-low mass, but mid and LF power-handling is about zero. Replacing the diaphragm material with stretched plastic films (in an attempt to extend LF bandwidth) merely results in degraded impulse response, defeating the reason for using a ribbon in the first place.
Horns and compression drivers have low distortion and the greatest headroom of all, but the extended bandwidth (on either end) just isn't there. Improving pulse response is the biggest challenge but not impossible - provided bandwidth limits are respected. Efforts to "extend" bandwidth result in gross increases in distortion and many additional resonances.
Big direct-radiators move a lot of air, work well in arrays at low frequencies, but store a lot of energy at the upper end of the range due to cone breakup and parasitic resonances from spiders and surrounds. This can be masked by skillful diaphragm damping but will still be evident in the time-decay signature, as well as increase in distortion. This is where low-pass filters come in handy - to restrict the energy going into those resonances, and keep distortion down.
Compression drivers were originally designed for movie theaters with 8 kHz bandwidth soundtracks, a minimum of equalization, and moderate-sized vacuum-tube power amplifiers with what we now call soft clipping (the 20 dB of feedback of the Williamson amplifier came 2 years after the 288 was designed). Hard clipping didn't really arrive until the Crown DC300 starting seeing professional use - and compression drivers started to fail in large numbers.
There is even a brief mention towards the end of the Lansing/Hilliard paper that the 515 and 288 drivers give their optimum performance with an amplifier output impedance equal to the driver impedance (unity coupling) - a detail that's been lost in the shuffle once high-feedback amplifiers became standard.
More than sixty years after the 288 was introduced, I find it significant that the most heroic efforts, using the most exotic diaphragms and phase plugs, are only one octave better than the 288. That says a lot about the HF limitations of compression driver technology, because there certainly has no been shortage of time, money, and talent thrown at it. If the best minds in the industry have been "working the problem" for that length of time, I would gently suggest that the HF limitations of compression drivers are not going away any time soon. The technology is trying to tell us something, if we will but listen.
This is why I was posting earlier about finding out the underlying limitations of the technology, and using the (inherent) limitations to guide the design process, instead of vainly hoping for some kind of breakthrough. Breakthroughs in audio technology are actually very rare, and almost always come with a new set of limitations of their own.
My rather simple-minded process is to look at each technology, and find what it is good at, and not so good at. Ribbons are superb at pulse response and ultra-low mass, but mid and LF power-handling is about zero. Replacing the diaphragm material with stretched plastic films (in an attempt to extend LF bandwidth) merely results in degraded impulse response, defeating the reason for using a ribbon in the first place.
Horns and compression drivers have low distortion and the greatest headroom of all, but the extended bandwidth (on either end) just isn't there. Improving pulse response is the biggest challenge but not impossible - provided bandwidth limits are respected. Efforts to "extend" bandwidth result in gross increases in distortion and many additional resonances.
Big direct-radiators move a lot of air, work well in arrays at low frequencies, but store a lot of energy at the upper end of the range due to cone breakup and parasitic resonances from spiders and surrounds. This can be masked by skillful diaphragm damping but will still be evident in the time-decay signature, as well as increase in distortion. This is where low-pass filters come in handy - to restrict the energy going into those resonances, and keep distortion down.
Lynn Olson said:
Lynn,
I think a lot of ppl here appreciate and respect your insight into the industry, and we all learned a great deal from your musings.
However, having in mind the original stated purpose of this thread -- and keeping this a productive dialog -- some (or _any_) progress in your construction more than one year after this thread is started would keep important contributors (like Magnetar, Michael and Salas) interested and involved.
Just $0.02.
Best regards, keep this coming.
Florian
Post #3401 was a superb summary of the previous 3400 posts; thanks, Lynn. You have provided a simple yet slam-dunk philosophy of WHAT you want to build, and WHY you choose to do it that way. While some may disagree, opinions never end.
I was expecting this process to take a number of years; we all would do well to exercise that difficult virtue called patience. The rewards will pay off, I believe.
Now if I could just convince you to start a similar thread on your Karna amp...........so many questions, so little time.
I was expecting this process to take a number of years; we all would do well to exercise that difficult virtue called patience. The rewards will pay off, I believe.
Now if I could just convince you to start a similar thread on your Karna amp...........so many questions, so little time.
Lynn Olson said:
I think that a philosophical evaluation of what we aim for with what we really have in our hands, using the benefit of hindsight, is clearly hitting the end transducer nail on its head. It cant be a hasty practical project by definition, but a paradigm of calm and right thinking. It has roots in your conscience vs psychology clarifications. Your end product will be an example of approach and we will all be happy to see it and hear it functioning most possibly via copy in some future audio fest. Knowing it first hand would be like a handshake with Lynn. Thanks for your contribution so far, its healthy bcs its open source and free minded. Something that is rare these days and has strong classical value. Wishing you health and potency so to practically make something relatively soon.
Hello,
For those who want to read the paper:
http://www.classicaudio.ru/articles/altec.pdf
see page 6:
"In the past it has been customary to adjust the amplifier output impedance to a value of approximately one-half to one-third of the average loudspeaker impedance. Improved performance can be obtained with the new loudspaker when the amplifier and loudspeaker impedance are approximately equal."
Best regards from Paris, France
Jean-Michel Le Cléac'h
(user of high output impedance amplifiers)
For those who want to read the paper:
http://www.classicaudio.ru/articles/altec.pdf
see page 6:
"In the past it has been customary to adjust the amplifier output impedance to a value of approximately one-half to one-third of the average loudspeaker impedance. Improved performance can be obtained with the new loudspaker when the amplifier and loudspeaker impedance are approximately equal."
Best regards from Paris, France
Jean-Michel Le Cléac'h
(user of high output impedance amplifiers)
Lynn Olson said:
Enough talk, let's see some REAL progress!
Methinks you're overplanning it- easy to do when you're competent and confident. But there's plenty you won't know until you start going through the mockup phase.
There's a photography adage (paraphrased):
"Most people take a few pictures. Good photographers take a lot of pictures. Professionals take a ton of pictures."
The idea being that the excellent result is as much the product of many tries as it is higher skill level. Certainly all of us speakerbuilders have been surprised at times by trying variations on a theme. Some things you just don't know until it's physical.
Some other forums have an emoticon that is a bunch of smilies holding signs that read: "This thread is worthless without pics"
Pretty much where this thread has gone. It's too long and cumbersome to 'jump into' and it has no build content to allow someone skimming the thread to pick up on where the project stands.
And I like a lot of your design philosophies, lynn. I'm sure it'll be a great speaker.
Methinks you're overplanning it- easy to do when you're competent and confident. But there's plenty you won't know until you start going through the mockup phase.
There's a photography adage (paraphrased):
"Most people take a few pictures. Good photographers take a lot of pictures. Professionals take a ton of pictures."
The idea being that the excellent result is as much the product of many tries as it is higher skill level. Certainly all of us speakerbuilders have been surprised at times by trying variations on a theme. Some things you just don't know until it's physical.
Some other forums have an emoticon that is a bunch of smilies holding signs that read: "This thread is worthless without pics"
Pretty much where this thread has gone. It's too long and cumbersome to 'jump into' and it has no build content to allow someone skimming the thread to pick up on where the project stands.
And I like a lot of your design philosophies, lynn. I'm sure it'll be a great speaker.
Lynn Olson said:
Hi Lynn,
For this driver the 2.83V sensitivity is 98.6dB and "no" is 3.61%. Is this what you needed or do you need some measured curves?
I will have to check with to see which Altec driver the cone was modeled after.
The magnets in these are dual 1" thick x 6 1/8" diameter Ceramic 5. We may plan on doing a neo motor in the future to get weight down for the pro audio usage. We also plan on some Alnico motors for our guitar and bass guitar drivers as most believe the Alnico has it's own "sound", but in the case of these drivers we wanted drivers that are as neutral as possible.
The driver is an overhung although different from most overhung drivers. It has a 3/4" gap plate with 1" long coil so the overhang is very short. The BL curve however is very broad. You can see the details on the motor design here:
http://www.aespeakers.com/Lambda001-1.php
The TD15's are all $279 each.
I believe the region you're looking at is the 70-200hz region, not 700-2KHz. My room is quite small so I have to time gate and measure only above about 500Hz to get very accurate. When measuring those I was mainly concerned with the response up higher. That curve is for just the TD15M's without the BMS coax. The sharp dip you see at 5KHz is an issue caused by the 2 15" drivers at 25" center to center distance. I'll try to take some measurements outside of a single TD15M if the wind dies down soon.
John