Love that the gap is translated as "cave" and thinking of it like that put a smile on my face. But since it's hard to see in there, it is sorta like a cave.
Yes, the idea of using winding numbers to flatten out the force is also implemented in the Purifi drivers. If you have their driver in hand and look into the cave (lol) with a good light, then you can see that the coil layers are not the same and there are gaps. It certainly evens out the inductance. But most people don't realise the point I want to make: Since only about 30% of the length of the coil is inside the cave or gap, that means that near 70% of the coil is not in the gap. In effect this part of the coil is not 'active' unless we have pistonic movement. The fact is that at all times the 70% is behaving as external wire wound resistors. They are outside the gap. So visually we can see that the idea of the amplifier's impedance is so low (and hence damping factor nominally), that means the amplifier is never a voltage source. And in fact it is the entire 100% of the DC resistance of the coil that stops that.
With dynamic speakers we cannot be talking about amplifier being voltage sources.
The real impedance is the coil resistance plus that of the amplifier, not to mention cable and so on. The amplifier cannot see more than one impedance value at any one frequency.
The guy who invented the term 'damping factor' took it back. His name is F. Langford-Smith. But he reversed himself because he realised that the driver's coil resistance would need to be zero Ohm. True story.
Suspensions? Yes, another important topic too. But I have come across bass drivers and swept sub-100Hz frequencies and heard noisy suspensions.
There is a potential issue when you have a class AB amplifier that uses feedback. I don't want to go on an anti-feedback rant because that is not me. I just want to understand what feedback does. For example, feedback is voltage feedback. You are effectively made the output, not an input in itself, but connected to an input that goes back to the input stage. The interface with the speaker is unlike any other interface. Using feedback and you make the amplifier go in the direction of being a voltage source. There is a guy called Hans van Maanen who has written EAS articles, and he designs Class AB amplifiers used in his active speaker design. They are low feedback and he does something else that you would never think you would do with active speakers, but he add passive filters that EQ the current of all three drivers. The reason is simple, what he wants is that when the amp transitions from Class A to Class B, he wanted the current to be in sync with the voltage. Even with relatively low current phase angles, he claims to have captured bursts of high order odd harmonic distortion. Even a small amount of that will be very obvious at mid-frequencies.
I suppose what I am saying is that the idea that some amplifiers are voltage sources (but not when connected to dynamic drivers) or current sources, I believe they are missing the point. All amplifiers are current delivery devices. It's how the current behaves when the amplifier is connected to speakers, that is what determines the sound of the amplifier. Modify that behaviour and you get lower distortion, and this is what Hans van Maanen is doing. He is reducing distortion because he has figured out something, that current matters. The first French video also fits into that picture. The Elsinore design (and others) also fits into that picture. But many are reluctant, but I believe in time this will get increasingly better understood. Maybe also lead to better amplifier design.
Here is Hans van Maanen's graphic:
Next post I am thinking about posting something I showed Menno Vanderveen before Covid struck. Menno is famous for his audio transformers. I subsequently met up with him in Holland. I asked him this question: "Menno, I am certain that we are listening to the current of the amplifier, not it voltage." I had prepared that question very deliberately and then just shut my mouth to see what he would say. "Yes, and the good news is that we should be able to prove it." I think I now can. Part of the puzzle will be in my next post, but near midnight here.
Yes, as for the acoustic source configuration.
I believe that when Jerome mentions controlling the speaker driver with feedback, this was regarding motional feedback via a servo attached to the cone.
Yes but two methods first is control motion with a sensor and the second is current sensing of the speaker, look like current drive.
We enter the beautiful world of electronic operations and DSP...
I am just going to post and see what comments we draw:
Three frequencies are chosen and compared when driven from a voltage source (actually measured) and the current and simulated in software.
42 Hertz: This is the peak sealed box resonance and hence Em motional impedance.
3KHz: Chosen as a critical frequency, it is said by some researchers that we can hear 0.01% distortion here and is dominated Ei inductive impedance..
20KHz: Basically chosen as it is the nominal limit (supposedly) of the audio bandwidth.
First the main graph and note the difference in dB-SPL at those three frequencies. Keep in mind that the impedance is a real measured impedance, not made up.
Again we see that Em and Ei are in series, and in turn with the Re DC Resistance:
Some number crunching, not too hard:
At those three frequencies, the change in dB-SPL between voltage source and current source, it is consistent with current.
Notably, the higher output, even at 42 Hertz, of the current source, that the 1A @6V is maintained constant across the board. The output with a voltage source, the current is freely able to change and since the impedance is always above the DC resistance of the voice coil. At 42 Hertz we may ascribe the peak in output as lack of damping (the response is actually limited in both by the mechanical Q damping). Hence damping can also be described in terms of current changing with frequency. But that is just an interesting side issue here. I am much more interested in what happens at 3KHz.
Note also that I have coined a variable called Vre as useful.
Let us see what others say. To me it says a number of things, one is that it point to a distortion mechanism where 3KHz would obviously be of interest as distortion is mighty obvious there.
Three frequencies are chosen and compared when driven from a voltage source (actually measured) and the current and simulated in software.
42 Hertz: This is the peak sealed box resonance and hence Em motional impedance.
3KHz: Chosen as a critical frequency, it is said by some researchers that we can hear 0.01% distortion here and is dominated Ei inductive impedance..
20KHz: Basically chosen as it is the nominal limit (supposedly) of the audio bandwidth.
First the main graph and note the difference in dB-SPL at those three frequencies. Keep in mind that the impedance is a real measured impedance, not made up.
Again we see that Em and Ei are in series, and in turn with the Re DC Resistance:
Some number crunching, not too hard:
At those three frequencies, the change in dB-SPL between voltage source and current source, it is consistent with current.
Notably, the higher output, even at 42 Hertz, of the current source, that the 1A @6V is maintained constant across the board. The output with a voltage source, the current is freely able to change and since the impedance is always above the DC resistance of the voice coil. At 42 Hertz we may ascribe the peak in output as lack of damping (the response is actually limited in both by the mechanical Q damping). Hence damping can also be described in terms of current changing with frequency. But that is just an interesting side issue here. I am much more interested in what happens at 3KHz.
Note also that I have coined a variable called Vre as useful.
Let us see what others say. To me it says a number of things, one is that it point to a distortion mechanism where 3KHz would obviously be of interest as distortion is mighty obvious there.
I wonder why you have so much power dissipation at 3kHz ?
Generally you have a high power demand in the low frequency.
Do you have a cone breakup at this frequency ? But I don't know how much it adds to the power demand. The cone breakup generates distortion.
This article below about cone breakup and EMF is very interesting 🙂
Speaker Break-Up and how to reduce his distortion
Second thing is you show when you drive by voltage, the current of the amplifier should not have limits.
All,
Today I had two audiophile friends over to listen to the new ULD Elsinores. They are both very familiar with the sound of my MFC Elsinores, having been over a number of times to listen to my various amplifier builds, system upgrades, etc. One is a doctor and pianist, who has some serious money in his system (Tidal Contriva speakers, Dave Berning PP845 ZOTL amps, among other high end equipment), and who is also is a dedicated tube audio DIYer, having built countless amplifiers over the years in pursuit of great sound. My other friend has a reputation in our local audio community as having the finest ears (he is informally nicknamed "Bat Ears"), someone who has great aural memory, and can quickly identify and characterize the impact of changes in the sound of the system (invaluable for tube and cable changes), even if the changes were made over the course of time (i.e., between visits). As a side note, he has been a fan of the sound of my MFC Elsinores in combination with the 211 Class A2 SET monoblocks that I designed and built for a number of local audio friends (including for him). Just establishing their bona fides.
We listened using three different amplifiers, spending most of the time using my 211 SET amps using the Chord DAVE DAC as a digital preamp. The other amps included a dual mono PASS FW DIY F4 driven by an 12SN7 Aikido preamp, and a KT66 Williamson amps, also driven by the Chord DAVE. We only listened to the ULDs, all comparisons were based on long time exposure and aural memory of the MFC Elsinores.
Both were tremendously impressed with the ULD Elsinores and felt that the speakers were a substantial improvement in speed and clarity over the MFCs. My pianist friend is very focused on transient reproduction, both the leading edge of the attack as well as the decay. He felt the MFCs were always a bit lacking in this area but the ULDs were quite good. He also felt the speakers were very revealing of recording quality, allowing one to hear the faults of poor engineering, but well engineered recordings sounding excellent. This is exactly what you want in a speaker, to be truthful to the material. Another major observation was how clean the sound is. My "good ears" friend has a pair of DIY compact 2-way horn speakers using the same Purifi driver as on the ULDs, so he is familiar with how this driver sounds and notes that these have that same clarity and lack of audible distortion. Both felt the ULDs seem to have much better HF extension and definition over their recollection of the MFCs.
Soundstaging was excellent, 3 dimensional, and completely detached from the speakers, which were only a visual presence but never locatable aurally.
Low frequency performance is cleaner and tighter than the MFC's, but does not have quite the same extension below 40 Hz based on the same source material heard previously. Some of that may be the drivers still need to loosen up more. I note that the LF response modeling I did a few weeks ago indicated a bit of a bass rise below 100 Hz with the SB Acoustics drivers whereas the Purifi drivers did not have this rise given the same enclosure parameters. I think this results in a bit more "fullness" to the sound of the MFC compared with the ULD. This fullness can be really appealing to some musical material, bringing a bit more "weight", but also a bit of mushiness (not as well damped?). My 211 SET amps sound tighter and better control on the bottom on the ULD Elsinores (not as mushy).
A switch to the PASS FW F4 confirmed that the 211 SET amps actually did control the ULD fairly well in the bass, the improvement on the bottom with a solid state amp was not as pronounced as one might expect compared to a SET amp, though the F4 did dig a little deeper. It basically confirmed that alot of the characteristic "SET" LF sound I was getting on the MFC Elsinores was really more an attribute of the speakers rather than the amp. Interesting. The F4 sounded really good, though I need to work a bit on the preamp (tube and output cap rolling) to get the best performance. This has the potential of being a fantastic match.
Finally my KT66 Williamson amps (using Heyboer reproduction Peerless output transformers), probably sounded the best of all the amps on the new ULD's in the session today. It really was a musical combination that was involving (sit back and enjoy the music). It also did an outstanding job controlling the bottom end.
Basically, the only (very mild) criticism of the ULDs today was a sin of omission, i.e., the bottom octave, and I am starting to look into adding a subwoofer active below 40Hz to fill this in better. This is an easy fix, once well integrated. I understand Joe did something similar as I recall.
Anyway, I am definitely a "happy camper", and this was a project that was well worth the expense and effort to undertake.
David
Today I had two audiophile friends over to listen to the new ULD Elsinores. They are both very familiar with the sound of my MFC Elsinores, having been over a number of times to listen to my various amplifier builds, system upgrades, etc. One is a doctor and pianist, who has some serious money in his system (Tidal Contriva speakers, Dave Berning PP845 ZOTL amps, among other high end equipment), and who is also is a dedicated tube audio DIYer, having built countless amplifiers over the years in pursuit of great sound. My other friend has a reputation in our local audio community as having the finest ears (he is informally nicknamed "Bat Ears"), someone who has great aural memory, and can quickly identify and characterize the impact of changes in the sound of the system (invaluable for tube and cable changes), even if the changes were made over the course of time (i.e., between visits). As a side note, he has been a fan of the sound of my MFC Elsinores in combination with the 211 Class A2 SET monoblocks that I designed and built for a number of local audio friends (including for him). Just establishing their bona fides.
We listened using three different amplifiers, spending most of the time using my 211 SET amps using the Chord DAVE DAC as a digital preamp. The other amps included a dual mono PASS FW DIY F4 driven by an 12SN7 Aikido preamp, and a KT66 Williamson amps, also driven by the Chord DAVE. We only listened to the ULDs, all comparisons were based on long time exposure and aural memory of the MFC Elsinores.
Both were tremendously impressed with the ULD Elsinores and felt that the speakers were a substantial improvement in speed and clarity over the MFCs. My pianist friend is very focused on transient reproduction, both the leading edge of the attack as well as the decay. He felt the MFCs were always a bit lacking in this area but the ULDs were quite good. He also felt the speakers were very revealing of recording quality, allowing one to hear the faults of poor engineering, but well engineered recordings sounding excellent. This is exactly what you want in a speaker, to be truthful to the material. Another major observation was how clean the sound is. My "good ears" friend has a pair of DIY compact 2-way horn speakers using the same Purifi driver as on the ULDs, so he is familiar with how this driver sounds and notes that these have that same clarity and lack of audible distortion. Both felt the ULDs seem to have much better HF extension and definition over their recollection of the MFCs.
Soundstaging was excellent, 3 dimensional, and completely detached from the speakers, which were only a visual presence but never locatable aurally.
Low frequency performance is cleaner and tighter than the MFC's, but does not have quite the same extension below 40 Hz based on the same source material heard previously. Some of that may be the drivers still need to loosen up more. I note that the LF response modeling I did a few weeks ago indicated a bit of a bass rise below 100 Hz with the SB Acoustics drivers whereas the Purifi drivers did not have this rise given the same enclosure parameters. I think this results in a bit more "fullness" to the sound of the MFC compared with the ULD. This fullness can be really appealing to some musical material, bringing a bit more "weight", but also a bit of mushiness (not as well damped?). My 211 SET amps sound tighter and better control on the bottom on the ULD Elsinores (not as mushy).
A switch to the PASS FW F4 confirmed that the 211 SET amps actually did control the ULD fairly well in the bass, the improvement on the bottom with a solid state amp was not as pronounced as one might expect compared to a SET amp, though the F4 did dig a little deeper. It basically confirmed that alot of the characteristic "SET" LF sound I was getting on the MFC Elsinores was really more an attribute of the speakers rather than the amp. Interesting. The F4 sounded really good, though I need to work a bit on the preamp (tube and output cap rolling) to get the best performance. This has the potential of being a fantastic match.
Finally my KT66 Williamson amps (using Heyboer reproduction Peerless output transformers), probably sounded the best of all the amps on the new ULD's in the session today. It really was a musical combination that was involving (sit back and enjoy the music). It also did an outstanding job controlling the bottom end.
Basically, the only (very mild) criticism of the ULDs today was a sin of omission, i.e., the bottom octave, and I am starting to look into adding a subwoofer active below 40Hz to fill this in better. This is an easy fix, once well integrated. I understand Joe did something similar as I recall.
Anyway, I am definitely a "happy camper", and this was a project that was well worth the expense and effort to undertake.
David
Unfortunately this article contains several errors. It appears to encapsulate a lack of understanding of breakup.
Cone breakup does not produce distortion, though it is responsible for various other issues.
Interesting I always read this assertion about hard cone, you have a rise of distortion at least at F breakup / 3. Example with SEAS Magnesium driver W22 breakup 4.5kHz rise of distortion 1.5kHz. And for soft cone I don't know because breakup is less visible but really present. It is not seems false at breakup frequency you can't have a good sound reproduction because the cone is ringing. I need some enlightment here...
Re:'transient reproduction, both the leading edge of the attack as well as the decay. He felt the MFCs were always a bit lacking in this area but the ULDs were quite good' - interesting to hear, confirms my experience with MFC drivers;
re:'does not have quite the same extension below 40 Hz' - I wonder if a combo would give the best of both worlds, MFC for the bottom two woofers, ULD for the upper pair?
re:'does not have quite the same extension below 40 Hz' - I wonder if a combo would give the best of both worlds, MFC for the bottom two woofers, ULD for the upper pair?
That's interesting...can you hear 0.01% distortion added to a 3kHz signal? Also, is that pure HD2 or pure HD3, or a combination of both? It would be interesting to create synthetic signals with the appropriate level of distortion added to see how they sound.
After creating a 3kHz signal and adding a 6kHz signal at the 0.01% level to it, I can't hear the presence of the synthesized 0.01% HD2 in the combined signal. I suppose that's not all that surprising, as 0.01% equates to a level of −80dB re the fundamental 3kHz signal.
@Joe Rasmussen, do you happen to have a reference for that 0.01% HD figure?
The distortion stays the same.
The harmonic can get manipulated to be more quiet, more loud or just the same. This can happen anywhere the response is not flat, not just at breakup... but breakup is unfairly singled out because it can have a big peak over a small angle which looks like a way to make a point.
Breakup also has a lot of dips, why don't people talk about the apparent reduction of measured distortion at breakup? The on-axis peak is not representative of the power after you measure on all axes and average them.
Better still, eliminate the breakup region completely. You'll benefit by freedom from a range of breakup problems 😉
Quickly read that article on breakup. It is an interesting topic but also misunderstood. Maybe one bit of required read is getting Martin Collom's tome "High Performance Loudspeakers" there are heaps of pages from 22 onward. Diaphragms and domes do all kind of bending modes. They don't stay pistonic. For example soft dome tweeters the centre of the dome becomes stationary (ideally) and should behave like a ring radiator.
I think I understand what Allen is saying here. I hope he doesn't mind me having a bit of a say on the topic, but when looking at drivers this is a big topic to me, when using 1st/low order low pass filters.
Breakup and resonance often gets confused and conflated as same or similar. Now please note, what follows are my observations and not meant to be the last word.
I will deal with MidBass drivers. You get to a certain frequency where the cone must flex or stay in tact. Mostly it flexes and after that, at some point, you may then see a droop and then a rise in response and some think that this rise is a breakup. Especially low inductor drivers (good) can have that rise, and it is most obvious on axis, this is actually a horn effect. Look at the cone shape, does it not look like a horn?
[Above is our MFC polycone and SB's measurement.]
See that "X" and I see one behaviour below and above that frequency. Note that dip? Very common, especially as this driver has very low inductance. The cone may be pistonic up to 1KHz or slightly above, but that kink is not 6dB down means that the cone has already started to flex and bend modes are happening (in theory the radiating area should reduce). I think this is what Allen says that it does not mean distortion even if the cone is not longer pistonic. Or else by that kink the response would 6dB down (Allen knows about Roy Allison who said a totally stiff cone would roll off 1st order). So in the MFC cone I see that it is predictable up to that 3KHz kink and ideally it should roll of controlled above 3KHz.
But it does not and many now assumes what they see above as breakups and resonances in the cone. That is not necessarily the case. Something else could well be happening...
Then on axis Blue we see a [irregular] horn effect. It has more to do with the shape of the cone looking like a horn and that comes into play above the kink, one cone behaviour below 3KHz and a different one above it. High inductance drivers may look smoother and 'less breakup' and this is misleading.
The first person I saw refer to this horn effect was Esa Merilainen in his "Current-Drive" book. But look 30° off axis and with good drivers, the horn effect can almost disappear. Would that happen if we had a severe resonance causing severe distortion? Then look at the CSD plot (waterfall) and is there ridge like resonances showing up? No, not always.
I discussed this with Lynn Olson and we agreed that above 3KHz or so, it is the cone shape that can dominate the frequency response. I supplied samples of two identical driver with cone of identical shapes and yet very different cone material, and yet they had near identical responses.
I was Lynn who said that good drivers (low distortion, low inductance) can actually look bad, because good low inductance drivers don't hide it whereas bad drivers it can hide the issue. The peak may indicate a better driver (which is why I always look off axis and look closer.) The article does not mention this, as far as I could see. See a peak and it must be bad? But don't be too quick to judge a good driver as bad. But this happens. Look above, that flat impedance? That is an ultra-low inductance (pretty much SOA for a driver of this type. It looks worse than it really is. A good driver can look bad.
OK, others will have their opinions. But that is what I see and I have had many discussions with others.
BTW, I know that Troels Gravesen has gravitated towards lower and 1st order crossovers (he likes the sound). He first questioned that I used the NRX-1 (first version) and he felt I should have used the MFC polycone as more suitable. He had a point. Eventually we had to adopt the MFC when NRX-1 was discontinued. My point to Troels, that he should not have been so quick to dismiss the NRX-1 because I was looking and forgiving the driver that he was too quick to dismiss. They were both suitable.
Now that brings me to the NBAC aluminium cone to round this off.
This is a 'hard' cone. If you applied pink noise or MLS noise to aluminium cones, you will hear that metal or hard cone character. Now we are talking about something different again. Once again the cone will stay pistonic, but it will try to get up to a higher frequency and when it gives out, it is likely to cause that metallic sound. That is not good. Except that sometimes there can be an exception. It was what Troels said about the NBAC driver from his Clio setup with MLS signal, he did not hear that hard cone zing. I looked closely (as he did) and my measurement showed a fairly benign resonance at 7KHz. But it was hardly audible. I could have used an LCR tuned to near "Y" 7KHz to deal with it, and I did. But I ended up not using it and choosing a higher value inductor to further suppress it 1st order, and was happy to accept what I saw as a bit of a BBC dip in the response, which I don't mind.
Allen might have his thoughts on the above (I am sure he does), but the above are my experiences rather than strict theory. Again, this went on a bit longer that I intended. Sorry.
PS: The TexTreme cone is also considered hard cone (aluminium, magnesium, ceramic etc), and it too avoids sounding like a hard cone, like the NBAC does. So maybe this will be a future trend that driver designers are coming to grips with this? The NBAC aluminium cone was designed by Ulrik Schmidt and the good outcome was not an accident. He has put these ridges radially out from the centre dome and he must have worked deliberately to get the right result. Are other hard cone designers also working on this? I hope so.
I think I understand what Allen is saying here. I hope he doesn't mind me having a bit of a say on the topic, but when looking at drivers this is a big topic to me, when using 1st/low order low pass filters.
Breakup and resonance often gets confused and conflated as same or similar. Now please note, what follows are my observations and not meant to be the last word.
I will deal with MidBass drivers. You get to a certain frequency where the cone must flex or stay in tact. Mostly it flexes and after that, at some point, you may then see a droop and then a rise in response and some think that this rise is a breakup. Especially low inductor drivers (good) can have that rise, and it is most obvious on axis, this is actually a horn effect. Look at the cone shape, does it not look like a horn?
[Above is our MFC polycone and SB's measurement.]
See that "X" and I see one behaviour below and above that frequency. Note that dip? Very common, especially as this driver has very low inductance. The cone may be pistonic up to 1KHz or slightly above, but that kink is not 6dB down means that the cone has already started to flex and bend modes are happening (in theory the radiating area should reduce). I think this is what Allen says that it does not mean distortion even if the cone is not longer pistonic. Or else by that kink the response would 6dB down (Allen knows about Roy Allison who said a totally stiff cone would roll off 1st order). So in the MFC cone I see that it is predictable up to that 3KHz kink and ideally it should roll of controlled above 3KHz.
But it does not and many now assumes what they see above as breakups and resonances in the cone. That is not necessarily the case. Something else could well be happening...
Then on axis Blue we see a [irregular] horn effect. It has more to do with the shape of the cone looking like a horn and that comes into play above the kink, one cone behaviour below 3KHz and a different one above it. High inductance drivers may look smoother and 'less breakup' and this is misleading.
The first person I saw refer to this horn effect was Esa Merilainen in his "Current-Drive" book. But look 30° off axis and with good drivers, the horn effect can almost disappear. Would that happen if we had a severe resonance causing severe distortion? Then look at the CSD plot (waterfall) and is there ridge like resonances showing up? No, not always.
I discussed this with Lynn Olson and we agreed that above 3KHz or so, it is the cone shape that can dominate the frequency response. I supplied samples of two identical driver with cone of identical shapes and yet very different cone material, and yet they had near identical responses.
I was Lynn who said that good drivers (low distortion, low inductance) can actually look bad, because good low inductance drivers don't hide it whereas bad drivers it can hide the issue. The peak may indicate a better driver (which is why I always look off axis and look closer.) The article does not mention this, as far as I could see. See a peak and it must be bad? But don't be too quick to judge a good driver as bad. But this happens. Look above, that flat impedance? That is an ultra-low inductance (pretty much SOA for a driver of this type. It looks worse than it really is. A good driver can look bad.
OK, others will have their opinions. But that is what I see and I have had many discussions with others.
BTW, I know that Troels Gravesen has gravitated towards lower and 1st order crossovers (he likes the sound). He first questioned that I used the NRX-1 (first version) and he felt I should have used the MFC polycone as more suitable. He had a point. Eventually we had to adopt the MFC when NRX-1 was discontinued. My point to Troels, that he should not have been so quick to dismiss the NRX-1 because I was looking and forgiving the driver that he was too quick to dismiss. They were both suitable.
Now that brings me to the NBAC aluminium cone to round this off.
This is a 'hard' cone. If you applied pink noise or MLS noise to aluminium cones, you will hear that metal or hard cone character. Now we are talking about something different again. Once again the cone will stay pistonic, but it will try to get up to a higher frequency and when it gives out, it is likely to cause that metallic sound. That is not good. Except that sometimes there can be an exception. It was what Troels said about the NBAC driver from his Clio setup with MLS signal, he did not hear that hard cone zing. I looked closely (as he did) and my measurement showed a fairly benign resonance at 7KHz. But it was hardly audible. I could have used an LCR tuned to near "Y" 7KHz to deal with it, and I did. But I ended up not using it and choosing a higher value inductor to further suppress it 1st order, and was happy to accept what I saw as a bit of a BBC dip in the response, which I don't mind.
Allen might have his thoughts on the above (I am sure he does), but the above are my experiences rather than strict theory. Again, this went on a bit longer that I intended. Sorry.
PS: The TexTreme cone is also considered hard cone (aluminium, magnesium, ceramic etc), and it too avoids sounding like a hard cone, like the NBAC does. So maybe this will be a future trend that driver designers are coming to grips with this? The NBAC aluminium cone was designed by Ulrik Schmidt and the good outcome was not an accident. He has put these ridges radially out from the centre dome and he must have worked deliberately to get the right result. Are other hard cone designers also working on this? I hope so.
Last edited:
Thanks for the effort to post that. Yes, you are right, every time we use a new driver (maybe no more) we get different set of compromises. But is the end result worth it? Well, the answer to that question is what matters.
Below is the ULD in my room with twin Subs (SVS SB3000 works very well in pairs and is the SVS range sweet spot, but must be in pairs IMO).
The room modes are obvious to me as I get the same ones with different speakers, but measured in the same spot in the room. This is quite credible and add that extra extension. BTW, I note that Earl Geddes suggest that extreme low LF should have a mild boost. Maybe the result above has a hint in that, but it may not be wrong to add around 5dB here. Geddes explains that this is because rooms are "leaky" at the very lowest frequencies. But it comes down to listening rather than taking his word for it. So if you get subs (pairs preferred), then explore what Geddes is saying and add around 5dB. Let your ears then be the last arbiter.
For some time I have heard from several people that we can definitely hear 0.05% distortion being reduced which means we can hear below that. But in this instance I had in mind what Earl Geddes has said publicly, that we can hear amplifier crossover distortion down to 0.01% and of course crossover distortion is going to be most audible around 2-3KHz, and he mentioned those frequencies. But types of distortions is of course a key aspect of these kinds of discussions.
I read all, Thank you ! I totally missed the horn effect. When you try to understand speakers nothing is simple.
Again Thank you for your very didactical answer 🙂
Isn't ringing just a word that describes a property of resonance? Resonance varies in several ways, it's a sign that response is higher or lower, or is rolling off.. but for example, should it matter that a vented box rings more than a sealed box?
Joe, would you care to elaborate what exactly you mean by the "horn effect", i.e. describe cause and effect a bit more in detail?
No for me Ringing is a consequence ? Why not a property with a frequency f ? when the cone is under an excitation. After the "horn effect", you have the "bell effect" LOL 🙂
That is a very large question because we use this term of resonance in a lot of domains 🙂
Not sure how I can really define resonance. I would say a resonance is when you can define in an unique frequency for a physic phenomenon in the frequency domain with a fourrier transform. And it is in newton mechanic. I digress a little with NMR, nuclear magnetic resonance, when the model to calculate the resonance frequency is a newtonian model and the theory is pure quantic physics.
Not sure about this definition, i didn't practice for a long time.
No but it is a matter of controls ? Just find the good compromise ?
Yes, exactly. Resonance has damping.
So when reading the link you posted, i saw...
It isn't so.