OK, I guess I could have started a new thread but realise that some of the more competent people on transducers inhabit this one. Over on the Planars and Exotics forum I have posted a thread called "Electrostatic AMT?" It describes an idea that I came up with some time ago but now realise that it requires input from others, particularly those versed in fluid dynamics. Are you there Mr Kreskovsky? In the, so far, limited response the soundness of the principle has been recognised but some cold water has been poured on the capabilities of a motor based on electrostatics. Any input would be greatly appreciated. AMT stands for Air Motion Transformer, the name Dr Oskar Heil gave his invention.
Keith
Keith
Last edited:
Not any combination of separately measured amplitude, nor group delay, nor phase, nor time responses can adequately describe what we perceive.
There needs to be an understanding that any crossover/loudspeaker having capacitive/inductive characteristics acting within the passband, affects the amplitude, phase and transient responses we hear during *music time*.
What happens to transients is quite different to what happens to steady sines, and thus reproduced music waveforms are constantly varying wrt the composite original waveform because music is never purely 'transient' or purely 'steady'.
(This is also where NFB amplifiers so often go wrong too, because not enough coincidentally occuring characteristics are studied simulataneously, and we end up 'hearing' the NFB response itself being superimposed upon the music waveform during music time.)
We can perceive all these problems, even with 'old' ears, and thus it MUST be our ears, not equipment, which determine the loudspeaker performance we *prefer*. I am following this thread to follow how Lynn's long experienced perception might shape design and development, because the graphs being shown CANNOT reveal what Lynn hears, and thus the loudspeaker *design* we are searching for.
I am not wishing to impose upon you Lynn, for of course response curves do give significant pointers towards characteristics and weaknesses, though in isolation not to how we might best 'enjoy' a finished design.
(Crossover inductors, and even voice coil inductance, change the shape of a transduced music waveform in time; this varying in direct relation to the music waveform itself at any moment during music time.)
There needs to be an understanding that any crossover/loudspeaker having capacitive/inductive characteristics acting within the passband, affects the amplitude, phase and transient responses we hear during *music time*.
What happens to transients is quite different to what happens to steady sines, and thus reproduced music waveforms are constantly varying wrt the composite original waveform because music is never purely 'transient' or purely 'steady'.
(This is also where NFB amplifiers so often go wrong too, because not enough coincidentally occuring characteristics are studied simulataneously, and we end up 'hearing' the NFB response itself being superimposed upon the music waveform during music time.)
We can perceive all these problems, even with 'old' ears, and thus it MUST be our ears, not equipment, which determine the loudspeaker performance we *prefer*. I am following this thread to follow how Lynn's long experienced perception might shape design and development, because the graphs being shown CANNOT reveal what Lynn hears, and thus the loudspeaker *design* we are searching for.
I am not wishing to impose upon you Lynn, for of course response curves do give significant pointers towards characteristics and weaknesses, though in isolation not to how we might best 'enjoy' a finished design.
(Crossover inductors, and even voice coil inductance, change the shape of a transduced music waveform in time; this varying in direct relation to the music waveform itself at any moment during music time.)
Last edited:
I would say just give it a run. Just keep in mind there are different tradeoffs among different designs.
News Tidbit
Just called Radian in Los Angeles at 714-288-8900, and got an answer about something I was curious about: what is the exit angle of the 1.4" format 745PB and 835PB compression drivers?
The answer, after a few minutes going through the engineering diagrams, was:
2.4 degrees - and the same for the 745PB and the 835PB.
They also confirmed that both models are the same except for the magnet size: the 745 has a 49.6 oz ceramic magnet with an external diameter of 6.5" that produces a 17,500 gauss field in the gap (resulting in 111 dB/metre efficiency), while the 835 has a 72 oz ceramic magnet with an external diameter of 7.5" that produces a 19,000 gauss field in the gap (resulting in 112 dB/metre efficiency). In functional terms, the gap of the 835 is pushed closer to saturation, thanks to the bigger magnet (Radian confirmed that the phase plug and gap assembly are the same in both models).
Although not the same as the Altec 288C that Bjorn Kolbrek measured at 8 degrees, fortunately, the difference is in a favorable direction: smaller rather than larger, so the wavefront coming out of a Radian 1.4" driver doesn't see a decrease in the slope of the horn walls when it enters the AH425 (which is optimized for 8 degrees).
I've become gradually convinced the possibility of mismatch between the internal exit angle of the compression driver and the entrance angle of the horn is why mixing-n-matching produces such unpredictable results; reflections in the first few inches of the wavefront as it expands through the horn produces big trouble by the time it leaves the horn. Of course, we have old-school radial horns that actually have intentional "pinches" in the throat area - equally troublesome from the standpoint of diffraction.
There was a request a few days ago (before the RMAF here in Denver) that I walk-through my goals again: no problem! It's pretty simple: I've wanted for a long time to have a truly high-efficiency loudspeaker (100 dB/metre or above) that has coloration levels approaching the Ariel, which in turn was based on the Quad ESL57 and BBC/Spendor/Harbeth loudspeakers. I was reflecting on this after my choral-director friend Gary Dahl visited the Harbeth room at the show and remarked how the Ariels are almost indistinguishable from them - and I told him that was no accident, the Ariel was specifically designed to sound as close as possible to that type of loudspeaker, while having 5~6 dB more efficiency.
It may seem absurd to think a Harbeth-alike with 100 dB or higher efficiency - well, I grant it's a bit of a stretch. But there's no reason high-efficiency speakers can't have lower levels of coloration.
Just called Radian in Los Angeles at 714-288-8900, and got an answer about something I was curious about: what is the exit angle of the 1.4" format 745PB and 835PB compression drivers?
The answer, after a few minutes going through the engineering diagrams, was:
2.4 degrees - and the same for the 745PB and the 835PB.
They also confirmed that both models are the same except for the magnet size: the 745 has a 49.6 oz ceramic magnet with an external diameter of 6.5" that produces a 17,500 gauss field in the gap (resulting in 111 dB/metre efficiency), while the 835 has a 72 oz ceramic magnet with an external diameter of 7.5" that produces a 19,000 gauss field in the gap (resulting in 112 dB/metre efficiency). In functional terms, the gap of the 835 is pushed closer to saturation, thanks to the bigger magnet (Radian confirmed that the phase plug and gap assembly are the same in both models).
Although not the same as the Altec 288C that Bjorn Kolbrek measured at 8 degrees, fortunately, the difference is in a favorable direction: smaller rather than larger, so the wavefront coming out of a Radian 1.4" driver doesn't see a decrease in the slope of the horn walls when it enters the AH425 (which is optimized for 8 degrees).
I've become gradually convinced the possibility of mismatch between the internal exit angle of the compression driver and the entrance angle of the horn is why mixing-n-matching produces such unpredictable results; reflections in the first few inches of the wavefront as it expands through the horn produces big trouble by the time it leaves the horn. Of course, we have old-school radial horns that actually have intentional "pinches" in the throat area - equally troublesome from the standpoint of diffraction.
There was a request a few days ago (before the RMAF here in Denver) that I walk-through my goals again: no problem! It's pretty simple: I've wanted for a long time to have a truly high-efficiency loudspeaker (100 dB/metre or above) that has coloration levels approaching the Ariel, which in turn was based on the Quad ESL57 and BBC/Spendor/Harbeth loudspeakers. I was reflecting on this after my choral-director friend Gary Dahl visited the Harbeth room at the show and remarked how the Ariels are almost indistinguishable from them - and I told him that was no accident, the Ariel was specifically designed to sound as close as possible to that type of loudspeaker, while having 5~6 dB more efficiency.
It may seem absurd to think a Harbeth-alike with 100 dB or higher efficiency - well, I grant it's a bit of a stretch. But there's no reason high-efficiency speakers can't have lower levels of coloration.
Last edited:
Member
Joined 2003
Lynn - will be interesting to know what you find with the Radian drivers. I have a pair from circa 1990 that would be the equivalent of the 745PB. OK drivers, but my Altec 291s sound much better. Maybe it's just a throat angle mismatch, or maybe something else. And maybe the modern versions are better - lot's of people like them.
Surprised you got the info - Radian have never been much help to me on the phone. Kinda disappointing. 2.4 degs seems right, my old Radians are almost straight.
Funny about the Harbeth. At T.H.E. Show in Vegas I went straight from the the Magnapan room to the Harbeth room. What a difference!
Surprised you got the info - Radian have never been much help to me on the phone. Kinda disappointing. 2.4 degs seems right, my old Radians are almost straight.
Funny about the Harbeth. At T.H.E. Show in Vegas I went straight from the the Magnapan room to the Harbeth room. What a difference!
The Super HL5. Very natural and silky-smooth sound, a refreshing change from the metallic, hard-edged sound of many systems at the 2009 RMAF. I would like to bring this kind of low-coloration sound to the world of high-efficiency speakers, if at all possible.
As for Radian vs Altec/GPA, I want to resolve the sample-variation issues in the 288 samples I have before messing around with the Radians. I've heard the 288's sound wonderful in other systems, so I expect these problems can be solved without too much trouble. There's a special magic - I can't pin it down, but it's there - with the classic large-format Altec drivers.
I also expect to contact a builder of the BFM cabinets in the next few months, and choose between the OmniTop 2x12, OmniTop 1x15, or DR280 cabinets. The one with the least need for equalization will probably be the one I choose. If I can get a 105 dB/metre system to be reasonably flat without too much EQ, I'll be a happy camper.
Last edited:
It was quite the show while it lasted...
http://www.diyaudio.com/forums/lounge/153403-flying-saucer-helium-balloon-new-post.html
dave
I´ve posted long ago in this thread. Because I know some here have expressed the importance of a smooth roll-off and large membane area, I wanted to point to a newish 10" Seas driver, the A26RE4.
There are two interesting aperiodic designs from Peter Comeau in World Audio Designs using this woofer and a large tweeter.
WD25A Design Pg1
Check also its elder floorstanding brother: http://wduk.worldomain.net/acatalog/pdf/WD25TReview.pdf
These designs root back in the Dynaco designs originating from "M. Stauning of SEAS Fabrikker".
There are two interesting aperiodic designs from Peter Comeau in World Audio Designs using this woofer and a large tweeter.
WD25A Design Pg1
Check also its elder floorstanding brother: http://wduk.worldomain.net/acatalog/pdf/WD25TReview.pdf
These designs root back in the Dynaco designs originating from "M. Stauning of SEAS Fabrikker".
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Thanks for pointing out the A26RE4. It really is a remarkable driver.
My buddy John brought a 2 way box to RMAF this year with the A26RE4 and the Revelator tweeter. Very, very good. It's amazing how "quiet" this driver is. Just music, not all the other junk we so often hear with speakers.
Those who were privileged to hear it were full of praise.
My buddy John brought a 2 way box to RMAF this year with the A26RE4 and the Revelator tweeter. Very, very good. It's amazing how "quiet" this driver is. Just music, not all the other junk we so often hear with speakers.
Those who were privileged to hear it were full of praise.
Hmmmm..... after the demo in the Maggie room, I found the Harbeths to be very colored. I could hear the wood and paper - big time. A very pleasant coloration, to be sure, but not neutral. Warm, woody. Or was it that the Maggies were colored in the cold direction? Or just a matter of taste?
Still, the Harbeth are a darn fine speaker. Very nice to listen to, very organic. State of the art in the format.
The Green Mountain speakers, that I've only heard very briefly at 2 shows, seem very clean, open and neutral to me. Do you know them?
Aha - great minds think alike!
OK, on a more serious note, yes, I've been thinking of the 414 option, comparing the Deltalite-II 2512 and Kappalite 3012HO specs to the Altec/GPA 414.
Looking at Augerpro's measurements (the GPA 414 is at the top), the compliance (Vas) is radically different, but the cone masses (a critical parameter) are pretty close, at 45 grams for the GPA 414, and 46.5 grams for the 3012HO. BL (another critical parameter) is 18.7 Wb/meter for the 414, and 14.7 Wb/meter for the 3012HO. Both drivers have a rising response above 500 Hz, a requirement for the horn. The Alnico 414 might need slightly more clearance for the magnet, and bolt-circle diameter might be a little different as well. The side-mounted vents (which double as cabinet handles) would probably have to be re-tuned for the 414, which would alter the level transition region where the short horn rolls off and the vent output takes over (180~250 Hz at a wild guess).
The 414 variation looks do-able to me, assuming you measure carefully and are willing to twiddle with the complex relationship between the horn and the vented box. I don't think BFM would be appreciative of hand-holding e-mail requests to re-design his cabinet, though - the forums on his site keep him plenty busy, from the looks of things. I am very appreciative he doesn't keep the details of his horns a big unmeasured secret, unlike so many other horn designs from the past and present. Just try and find frequency-response measurements of the Altec A5 and A7, the Klipschorn, Belle Klipsch, or La Scala, or the modern Avante-Garde Trio, for that matter.
I have the full set of BFM documents, and for the OmniTop 12 the optimal T/S specifications are: Fs of 45 Hz, Qts of 0.35, and Vas of 100 liters. The recommended drivers are the Eminence Basslite 2012 (lowest-cost option), Deltalite-II 3012, and the premium high-power options, the Kappalite 3012HO and Delta Pro 12. The passive 2nd-order lowpass filter for the woofer creates 2 dB of peaking in the 1.6 to 2 kHz range, in order to meet the 2 kHz crossover of the piezo array (which I will not be using).
The design of the DR280 curved horn is fascinating, and I can see why it is rated as a "difficult" build. The range of recommended drivers for the DR280 is more restrictive, with the Deltalite-II 2512 (fs=37 Hz, Qts=0.39, Vas=147 liters) being the first choice, and the Kappalite 3012HO as a premium option, with 2 dB higher midrange sensitivity than the 2512, and 450 watt power handling. (The comments in the BFM forums indicate the DR280 is best used with fairly strong active EQ, in the form of the well-known Behringer device.)
Ordering the plans from BFM is pretty inexpensive - and they include both dimensional callouts in the Word and PDF documents as well as separate Google Sketchup files so you can see what they look like in 3D. Just last night I was comparing the sizes and interior bracing arrangements of the OT15 (single) to the OT212 (dual). There are lots of subtle details in these plans that are interesting reading.
OK, on a more serious note, yes, I've been thinking of the 414 option, comparing the Deltalite-II 2512 and Kappalite 3012HO specs to the Altec/GPA 414.
Looking at Augerpro's measurements (the GPA 414 is at the top), the compliance (Vas) is radically different, but the cone masses (a critical parameter) are pretty close, at 45 grams for the GPA 414, and 46.5 grams for the 3012HO. BL (another critical parameter) is 18.7 Wb/meter for the 414, and 14.7 Wb/meter for the 3012HO. Both drivers have a rising response above 500 Hz, a requirement for the horn. The Alnico 414 might need slightly more clearance for the magnet, and bolt-circle diameter might be a little different as well. The side-mounted vents (which double as cabinet handles) would probably have to be re-tuned for the 414, which would alter the level transition region where the short horn rolls off and the vent output takes over (180~250 Hz at a wild guess).
The 414 variation looks do-able to me, assuming you measure carefully and are willing to twiddle with the complex relationship between the horn and the vented box. I don't think BFM would be appreciative of hand-holding e-mail requests to re-design his cabinet, though - the forums on his site keep him plenty busy, from the looks of things. I am very appreciative he doesn't keep the details of his horns a big unmeasured secret, unlike so many other horn designs from the past and present. Just try and find frequency-response measurements of the Altec A5 and A7, the Klipschorn, Belle Klipsch, or La Scala, or the modern Avante-Garde Trio, for that matter.
I have the full set of BFM documents, and for the OmniTop 12 the optimal T/S specifications are: Fs of 45 Hz, Qts of 0.35, and Vas of 100 liters. The recommended drivers are the Eminence Basslite 2012 (lowest-cost option), Deltalite-II 3012, and the premium high-power options, the Kappalite 3012HO and Delta Pro 12. The passive 2nd-order lowpass filter for the woofer creates 2 dB of peaking in the 1.6 to 2 kHz range, in order to meet the 2 kHz crossover of the piezo array (which I will not be using).
The design of the DR280 curved horn is fascinating, and I can see why it is rated as a "difficult" build. The range of recommended drivers for the DR280 is more restrictive, with the Deltalite-II 2512 (fs=37 Hz, Qts=0.39, Vas=147 liters) being the first choice, and the Kappalite 3012HO as a premium option, with 2 dB higher midrange sensitivity than the 2512, and 450 watt power handling. (The comments in the BFM forums indicate the DR280 is best used with fairly strong active EQ, in the form of the well-known Behringer device.)
Ordering the plans from BFM is pretty inexpensive - and they include both dimensional callouts in the Word and PDF documents as well as separate Google Sketchup files so you can see what they look like in 3D. Just last night I was comparing the sizes and interior bracing arrangements of the OT15 (single) to the OT212 (dual). There are lots of subtle details in these plans that are interesting reading.
Attachments
Last edited: