Paper cones can be build so that the cone flexes in a manner that makes the core/dome/dustcap able to play higher frequencies at the same time as the outer part of the cone is not. This makes a lot of paper cones able to play higher frequencies - on-axis - than hard cones. But they lose detail and pistonic movement during this flexing. So it sacrifices piston movement and clarity, to get higher frequencies, better self-damping features and smoother breakup-modes.
Last edited:
The Sabrina? Hmmm... I was impressed by its ability to produce natural voice that I have never heard before with a certain female singer. (I forgot the amplifier used then).
I agree... always like Wilsons midrange/tweeter combo. It's their bass performance I dont like - Sabrina's little farty ported woofer
Would you really put Wilson in the same league as the two I referred to? Then you are veeeery susceptible to marketing and commercial tricks, oh dear.
The part about better self-damping and smoother breakup is questionable. Paper doesn't excel on those qualities. Unless you mean better than e.g. metal. But paper actually performs worse than quite a few other materials. Paper is very cheap and easy to use though.
So.... you would say different kinds of fiber, like carbon is a better option?
I like mostly ceramics and metal - playing on a small Dayton RS125 right now and is very please with it's smooth and detailed character. together with a Seas DXT, it almost disappears in the sound image in front of me. Now I just need better midwoofer performance
No. I just disagree with the statement that paper would be the best, since all else is never equal. A well-designed driver with a paper cone can do a better job than a crappy metal or kevlar cone driver (some examples come to mind here...), but if the system it's in is poorly designed (I won't start about Wilson) from an electro-acoustic point of view, it still sucks.
Thanks... that makes sense. Because I have heard all kinds of speaker with a lot of different drivers and neither of them would excel because of a specific drivers choice.... even though that would make it easier
But FR wise, I do find harder cone drivers to be smoother and quite easy to handle with a modern DSP filter. My little Dayton RS125 midrange, requires only a low Q damping at around 2200hz to make it flat through its usable range. Further - it's breakup is around 10kHz - so even the 2 and 3 harmonics is well beyond my cross-over at 2kHz. My old Accuton midrange broke up at 4800hz... but still easy to handle with precise EQ, so that it still sound very nice and smooth. It seems like we can discuss technical details forever.... but with a few guidelines and maybe swapping drivers with friends.... we can find the right drivers with some trial and error
But FR wise, I do find harder cone drivers to be smoother and quite easy to handle with a modern DSP filter. My little Dayton RS125 midrange, requires only a low Q damping at around 2200hz to make it flat through its usable range. Further - it's breakup is around 10kHz - so even the 2 and 3 harmonics is well beyond my cross-over at 2kHz. My old Accuton midrange broke up at 4800hz... but still easy to handle with precise EQ, so that it still sound very nice and smooth. It seems like we can discuss technical details forever.... but with a few guidelines and maybe swapping drivers with friends.... we can find the right drivers with some trial and error
A while ago, I got curious about the break up on my Visaton 170. I skipped the notch filter in one speaker, also tried to switch off tweeters, listening only on the 170's.
I couldn't distinguish any difference with an x-over at 2kHz, 18db acoustical slope. Maybe in a domestic environment as mine, with not so high levels, it doesn't matter so much...
Diffraction and uneven response seems to be much more noticeable together with the room.
I couldn't distinguish any difference with an x-over at 2kHz, 18db acoustical slope. Maybe in a domestic environment as mine, with not so high levels, it doesn't matter so much...
Diffraction and uneven response seems to be much more noticeable together with the room.
Agreed. I found that building a midrange/tweeter box with the Seas DXT + a small 4-5" midrange, crossed around 2kHz in a rather narrow and/or beveled cabinet, made the best sound I ever had, when EQ'ed very flat with minimum effort and correctly designed front baffle.
Hi,
does it need a lot of parts in the filter to mzke the DXT flat?
I noticed indeed many testimonied about the fact than despite the waveguide, the front bafle needs to stay narrow or huge (I understand here something like a Sonus Faber Stradivari, or open bafle a la Troels Gravsen) ???
Any thoughs on this please ? Did you need to do complex cabinet shaping like Heissman does for instance ?
does it need a lot of parts in the filter to mzke the DXT flat?
I noticed indeed many testimonied about the fact than despite the waveguide, the front bafle needs to stay narrow or huge (I understand here something like a Sonus Faber Stradivari, or open bafle a la Troels Gravsen) ???
Any thoughs on this please ? Did you need to do complex cabinet shaping like Heissman does for instance ?
Hi,
without waveguide the baffle affects directivity / power response of the speaker. This is called the bafflestep Another thing that might affect power response is if some of your mid/bass drivers get directional below its low-pass filter.
I've got no further insight what is good and what is bad here, but you probably want these different sources of power response issues to help each other out rather than combine and make situation worse than either of them would alone. You've probably read it somewhere before that the bafflestep is usually handled with the crossover. Crossover frequency and baffle width are chosen so that adding some attenuation to the higher band(s) counteracts the "bafflestep".
Having bafflestep factored in the crossover is the easiest way to do it I believe. Not sure if it yields smoother power response than having bafflestep lower or higher than the crossover and then use additional filters to smooth it out. I've never done passive crossovers so never thought about this before. Narrow baffle makes sense if you think about diffracting edges as another sound source, just like the driver itself. If the edge is "touching" the driver the diffraction is indistinguishable from the direct sound (simplified).
Anyway, baffle width affects the power response. It is your waveguide when there is none.
edit: thinking more about it. With mid/bass driver, baffle as narrow as possible, puts the bafflestep where the driver starts beaming. This is about where you would want to crossover to the higher audio band anyway since both drivers have similar directivity (omni). Now the bafflestep is also about at the crossover frequency, which should be easy to deal with in the crossover? With tweeter there is no more low-pass filtering and there is already one bafflestep planned so you probably don't want another make another by narrowing the baffle for the tweeter? But, then there is diffraction from the tweeter since baffle edges are further away from the tweeter. Here you'd apply some rounding / slant around the tweeter to lessen diffraction. Conclusion, when no other issues need to be taken into consideration, baffle width would be chosen to the width of the mid driver. Use roundovers for the tweeter. Make crossover to the bafflestep. Many issues remain: where to put the bass driver if it is three way, depth and height limits for the enclosure etc.
without waveguide the baffle affects directivity / power response of the speaker. This is called the bafflestep
Another thing that might affect power response is if some of your mid/bass drivers get directional below its low-pass filter. I've got no further insight what is good and what is bad here, but you probably want these different sources of power response issues to help each other out rather than combine and make situation worse than either of them would alone. You've probably read it somewhere before that the bafflestep is usually handled with the crossover. Crossover frequency and baffle width are chosen so that adding some attenuation to the higher band(s) counteracts the "bafflestep".
Having bafflestep factored in the crossover is the easiest way to do it I believe. Not sure if it yields smoother power response than having bafflestep lower or higher than the crossover and then use additional filters to smooth it out. I've never done passive crossovers so never thought about this before. Narrow baffle makes sense if you think about diffracting edges as another sound source, just like the driver itself. If the edge is "touching" the driver the diffraction is indistinguishable from the direct sound (simplified).
Anyway, baffle width affects the power response. It is your waveguide when there is none.
edit: thinking more about it. With mid/bass driver, baffle as narrow as possible, puts the bafflestep where the driver starts beaming. This is about where you would want to crossover to the higher audio band anyway since both drivers have similar directivity (omni). Now the bafflestep is also about at the crossover frequency, which should be easy to deal with in the crossover? With tweeter there is no more low-pass filtering and there is already one bafflestep planned so you probably don't want another make another by narrowing the baffle for the tweeter? But, then there is diffraction from the tweeter since baffle edges are further away from the tweeter. Here you'd apply some rounding / slant around the tweeter to lessen diffraction. Conclusion, when no other issues need to be taken into consideration, baffle width would be chosen to the width of the mid driver. Use roundovers for the tweeter. Make crossover to the bafflestep. Many issues remain: where to put the bass driver if it is three way, depth and height limits for the enclosure etc.
Last edited:
The Heissmann is actually quite simpel. Its just at matter of keeping the baffle narrow around the tweeter - you can almost freely choose whether you want it to be done by bevels, roundings or other type of cuts. Its simply a matter of not having a too edgy or flat surface around the DXT. The DXT will send to much energy forward in the area around 2-3kHz because a 1" start to become omnidirectional here. And since the midrange starts to beam... we need a narrow baffle(the flat part) around the tweeter, to make them behave similar and create a smooth directivity.
I simply measure the tweeter and found the simplest PEQ that also made the tweeter nicely flat - but I use DSP. I have no clue how to do this with passive components... and I never will
Mine needed a PEQ at 2633hz Q=0,552 and -7,86dB + 6881,8hz Q=1,732 and -3,05dB.
Agreed with your disagreement. Paper is not and can't the best material for domes/cones.
Why? Paper is inherently damping. Internal damping would certainly be a nice property for a dome/cone. But this internal damping would also have to be uniform over the whole frequency range for a dome/cone to behave uniformely as well. And this might not be the case.
There are also doubts that any other damping material might show a linear internal damping behavior over the frequency range. And this hererogenous damping therefore implies an intrinsic flaw and comes along with compromises for it's use. Therefore a paper cone by itself will always be a more or less well archieved compromise right from the begin. Regardless of the geometry of the dome/cone.
You may try to archive a better behavior with a more or less magic mixture along with other and different damping materials. Or compound constructions. Or coatings. You also may use kevlar, carbon or bextrene instead of, or in addition to paper. Or Polywhatsoever. Same basic problem. You must have a well-defined, frequency linear damping behavior, and you may never ever get this perfectly with all these materials. If you are really knowing what you do, you certainly may quite probably approximate your goal of a well behaved damping over the frequency range. Which will result in a decently well behaving driver. But you might have done better altogether.
Or else, go to a cone/dome made of non-damped, and therefore resonant material. No problems then with this ill-behaved damping, what a freedom! Knowingly, you will have to use such cones/domes well below it's first resonance/breakup mode, at less than 1/3 of it's frequency of the first resonance. So building you driver as a designer, you only will have to try to shift the first cone breakup up as high as possible in the frequency range.This approach is straightforward and promises flawless results in terms of predictable cone behavior over the useful frequency range. From my experience, choosing max <= f_breakup / 4 is safe (and this is why I might forever be waiting for an affordable and really breathtaking tweeter with a first dime breakup frequency > 60kHz.)
If you want to stick with paper instead, you also will have to restrict the bandwidth of your paper dome/cone anyway and as well (like with a resonant dome/cone material, but for a different reason). You will have to restrict the bandwith to hold it within an acceptable range of non-linear damping within it's passband. That's at best a viable compromise and not very satisfying in terms of looking for the best possible. With resonant cones you try to control the nonlinearities far outside the passband instead.
You still want near-perfection along with paper? Yes, ok, go ahead, this will be possible if you redefine it's use
: In this latter never-give-up-logic you may end up with a paper coned driver if you restrict it's bandwidth to just one (!) single frequency to be used with. Heureka! No more damping deviations then over an annoying whole range of frequencies. Do that and be proud of yourself. And then you even may marketing this approach as your ingeniously unique and trademarked NBODC feature. Narrow Band Optimized Damping Control. Congratulations for your success. By the way - It might have a reason that in tweeters paper use has altogether vanished over time. Everybody may try his/her own educated guess on this.
Nota bene/Disclaimer: This all is a theory, for me a plausible one.
Last edited:
thanks guys,
I really hesitate in my project about the tweeter hence the questions that are a little off topics so I stop here questioning. My front bafle is 45 heigth and 28 cm width and 100 % edgy with 90° angle as it's a british like monitor. the ScanSpeak10F has its center at 11 cm from the nearest side, so not centered on the front bafle. Cut off around 500 hz, so 100 to 200 hz below the beginning of the bafle step drop... Here the culprit is the the 8" aluminium from ScanSpeak which has too much THD 5 around 1300 hz for a higher cut off with a LR12 (passive) electrical I chose.
The 10F will be crossed around 3k to 4k but most of the tweeter wave guide will be bigger than the 6.5 cm of the 10F radiating area, more around 10 cm diameter. The Seas DXT, Morel Cat368 also horned, and some AugerPro 3D print WG are appealing with their eleptical shape... or it it will be the SB26 CAC or Seas 22 TF/G... while I prefer to live without FerroFluid. description is for the story as it's off topic here but maybe how to mùarry a treble WG with the best midranges, héhé !
PatrickJohn Batman thread about bafle being a sort of horn, at least reflective shaping surface of the drivers power response is also helping with your answers... thanks for that
I really hesitate in my project about the tweeter hence the questions that are a little off topics so I stop here questioning. My front bafle is 45 heigth and 28 cm width and 100 % edgy with 90° angle as it's a british like monitor. the ScanSpeak10F has its center at 11 cm from the nearest side, so not centered on the front bafle. Cut off around 500 hz, so 100 to 200 hz below the beginning of the bafle step drop... Here the culprit is the the 8" aluminium from ScanSpeak which has too much THD 5 around 1300 hz for a higher cut off with a LR12 (passive) electrical I chose.
The 10F will be crossed around 3k to 4k but most of the tweeter wave guide will be bigger than the 6.5 cm of the 10F radiating area, more around 10 cm diameter. The Seas DXT, Morel Cat368 also horned, and some AugerPro 3D print WG are appealing with their eleptical shape... or it it will be the SB26 CAC or Seas 22 TF/G... while I prefer to live without FerroFluid. description is for the story as it's off topic here but maybe how to mùarry a treble WG with the best midranges, héhé !
PatrickJohn Batman thread about bafle being a sort of horn, at least reflective shaping surface of the drivers power response is also helping with your answers... thanks for that
Last edited:
well, some damped paper, paper mixed with something if you prefer are not bad for mids, look at the Audax Aerogel that are not peaky in their high ranges, though not the more linear in their usable range too...
Someone who heard also some 13 and 16 cm from Supravox that are 100% thin OCB paper like, sort of, don't care anymore of a dead flat power response... psycho acoustic is strange (I mean adaptive). Voxative and MarkAudio I don't know are said to have very good sounding result.
But I must admit, I'm more attracted at reading the datasheets of some drivers for selection towards Wavecores, Revelators and I wish SB Acoustic had a 4,5" ceramic that exhibits the very good same THD numbers of their awesome ceramic 6"... I personally like more small mids around 5", certainly because the soundstage is more liked by my ears... I dunno !
Edit : many thanks Pida, awesome usefull for my project.
Someone who heard also some 13 and 16 cm from Supravox that are 100% thin OCB paper like, sort of, don't care anymore of a dead flat power response... psycho acoustic is strange (I mean adaptive). Voxative and MarkAudio I don't know are said to have very good sounding result.
But I must admit, I'm more attracted at reading the datasheets of some drivers for selection towards Wavecores, Revelators and I wish SB Acoustic had a 4,5" ceramic that exhibits the very good same THD numbers of their awesome ceramic 6"... I personally like more small mids around 5", certainly because the soundstage is more liked by my ears... I dunno !
Edit : many thanks Pida, awesome usefull for my project.
Last edited:
Me? Susceptible to marketing tricks? You're mistaken there. I have my ears. When I heard a natural voice, it's natural voice to my ears regardless of prices, brands, graphs....
Other people OTOH, hear a Wilson Audio then said: "The frequency response is all over the place, how can this sound be good?" Trust me, these people will have a hard times in Klippel Test
My analogy regarding cook and meal is important too. Sound reproduction is very complex. The best midrange driver will have very little contribution in making a great sound system.
But if you listen to music everyday, you must be very happy because of it, then regardless of price I think you have a great sound system already
The guy at HifiCompass who is seing and measure a lot of drivers, I assume more than the average of us, is fond of the Illuminator Paper... which is in his everyday speaker if my memory serves me well enough. But by the hell, who want a so much horrible loudspeakers than the Wilson Audio are selling... awefull they are, but behind an giant acoustic black grill as width as the front wall, I could not live with that ! How one can have so huge bad tastes ! Should look at Sonus Faber elegant design a little for inspiration !
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.