Popart,
Notice that I mentioned time aligned "music" signals.
Notice also that some reflection related with the person will occur whether you use speakers or listen live. So this is a common factor that we can put aside. Room reflections will create other effects, and this also effects other speaker design aspects. If we can use the room in a way that in complements limitation in existing speakers, then it is good. This is also why subs are so popular.
Notice that I mentioned time aligned "music" signals.
I'm sure that lot's of people will say that this is not the goal. But I think each engineer can have his own set of goals that he wishes to gradually work towards. I respect that. I just like to focus on something that seems possible to some extent, but not accomplished yet.soongsc said:
Notice also that some reflection related with the person will occur whether you use speakers or listen live. So this is a common factor that we can put aside. Room reflections will create other effects, and this also effects other speaker design aspects. If we can use the room in a way that in complements limitation in existing speakers, then it is good. This is also why subs are so popular.
There doesn't seem to be much information on this resonance. I wonder if there would be a corresponding tell-tale blip on the electrical impedance as it would help me to draw a connection to this hole at 350Hz. I really don't want to set up for interferometry (well, OK, I really do
)Attachments
Here's data from a B&C 10PS26.
Near field frequency response:
Impedance:
Near field frequency response:
An externally hosted image should be here but it was not working when we last tested it.
Impedance:
An externally hosted image should be here but it was not working when we last tested it.
What a funny looking impedance curve! Why doesn't the impedance keep rising? And what's up with the vertical scale? Magnitude in dB's?
It seems to me that a mass coupled to a electromagnetic force best be done by a very low impedance coil. 8 ohms seem to be really high if one cares for speed in that respect. Anybody any insights on having speakers with really low impedance?
A higher impedance should be accompanied by a higher BL product for the same amount of electrical damping. Low impedance means thick wire, meaning less wire length fits in the coil gap and BL goes down. If you double the wire cross sectional area, wire length will halve, BL will halve, impedance will be 1/4 and you end up with the same electrical damping factor.
If you manage to obtain both a low impedance and high BL, you end up with a large electrical damping, meaning that there will be no bass when using a voltage amplifier. That's just the consequence of having a strong motor when you look at it in the frequency domain. The same happens when you reduce cone mass.
A low impedance and high BL is pursued in IPAL subwoofers, though for a different reason: it results in a higher efficiency. The lack of bass can be circumvented by using a current amplifier or by just applying EQ.
If you manage to obtain both a low impedance and high BL, you end up with a large electrical damping, meaning that there will be no bass when using a voltage amplifier. That's just the consequence of having a strong motor when you look at it in the frequency domain. The same happens when you reduce cone mass.
A low impedance and high BL is pursued in IPAL subwoofers, though for a different reason: it results in a higher efficiency. The lack of bass can be circumvented by using a current amplifier or by just applying EQ.
Last edited:
Good points. I'm guessing BL is less important than low Rdc concerning the accuracy of the translation of current to air pressure.
By damping you mean the damping of the speaker by the amplifier right? How can any speaker be damped electrically when the coil is in series with it's Rdc (8 ohms)? It seems there is hardly any actual damping possible this way, or the theoretical model I know isn't right. Could very well be, but how does it work then?
Not being able to comment on all your points (my bad;-), but I can sum up some strong points:
If you do the math and look for a low impedance coil with about the same dimensions it is possible to get a equivalent or higher efficiency, a lot lower impedance peak of Fs and I'm guessing a lot faster driver: inductance is minimal.
It does have some strong drawbacks of course, but it wouldn't need that much mechanical damping (lower Fs) and less voltage amplification of the signal. One would be able to make one with e.g.anodised aluminum foil, at about 50 uM thick.
Btw i never got this, but you could explain why a coil with e.g. a sub 1 ohm impedance that moves in a magnetic field has to have a high Fs? In my head this wire just moves a membrane, why peak significantly and create a resonance higher? Maybe we should feed it by current drive?
By damping you mean the damping of the speaker by the amplifier right? How can any speaker be damped electrically when the coil is in series with it's Rdc (8 ohms)? It seems there is hardly any actual damping possible this way, or the theoretical model I know isn't right. Could very well be, but how does it work then?
Not being able to comment on all your points (my bad;-), but I can sum up some strong points:
If you do the math and look for a low impedance coil with about the same dimensions it is possible to get a equivalent or higher efficiency, a lot lower impedance peak of Fs and I'm guessing a lot faster driver: inductance is minimal.
It does have some strong drawbacks of course, but it wouldn't need that much mechanical damping (lower Fs) and less voltage amplification of the signal. One would be able to make one with e.g.anodised aluminum foil, at about 50 uM thick.
Btw i never got this, but you could explain why a coil with e.g. a sub 1 ohm impedance that moves in a magnetic field has to have a high Fs? In my head this wire just moves a membrane, why peak significantly and create a resonance higher? Maybe we should feed it by current drive?
To correct myself, I meant 'damping ratio' instead of 'damping factor'.
It still does, see the equation for Qes: Thiele/Small parameters - Wikipedia
If you double the wire cross sectional area, wire length will halve, BL will halve, impedance will be 1/4x. Choose to fix input power to the same value. 1/4x impedance means 1/2x voltage and 2x current. BL was halved, so 2x current means that the force on the cone remains the same. Equal force means equal sound output power, so efficiency is equal too.
I might be misinterpreting what you say. Could you please clarify? Fs and Q (related to damping) are two different parameters.
When Re goes down and BL remains the same: fs remains the same, but Qes is lower. In this graph you can see the effect of different Q values: http://www.swarthmore.edu/NatSci/echeeve1/Ref/FilterBkgrnd/highpass1_2.gif
Last edited:
Damping ratio/factor, if this is right: Damping factor - Wikipedia then it follows that there hardly is any change in ratio, as the output impedance of the amplifier is many times lower than the Rdc of the speaker. I don't see how T/S parameters change this, please elaborate.
Regarding efficiency: all else being equal, efficiency is L×L divided by Re. Assuming the speakers are the same (magnet, air gap, etc), only with extreme low Re (<<1 ohm) the efficiency would be similar or higher.
None of it really matters though, the idea comes from the assumption that it's rather a soft way of accelerating a cone, if the Re is what it is these days.
I can not see how a mass with a resistance in series with the voicecoil be expected to have an excellent accuracy.
Regarding efficiency: all else being equal, efficiency is L×L divided by Re. Assuming the speakers are the same (magnet, air gap, etc), only with extreme low Re (<<1 ohm) the efficiency would be similar or higher.
None of it really matters though, the idea comes from the assumption that it's rather a soft way of accelerating a cone, if the Re is what it is these days.
I can not see how a mass with a resistance in series with the voicecoil be expected to have an excellent accuracy.
mterbekke said:
I might be misinterpreting what you say. Could you please clarify? Fs and Q (related to damping) are two different parameters.
I was referring to your original deduction that a lower Qes and same BL would result in a driver with a higher resonance frequency. Since the total Q of the system would be a result of the mechanical and electrical Q, my proposed solution to (partially perhaps) that change in Q was to use a suspension with less resistance, thereby lowering the resonance frequency. I'm trying how this relates to Q. Strangely it would need a higher Qms to get to the same Qts right? Hmm.
Btw, typing this on a phone and it's not the best way regarding layout and readability I'm afraid. Sorry for that.
Last edited by a moderator:
What imho is a good example of omitted quantification of a driver (or a speaker for that matter) and what might also drive perceived characteristics as dynamics etc is the interaction of the amplifier with the driver. These devices are of course tightly coupled. Let's take an example:
Damping factor
A clear example of how the amplifier and the speaker's impedance and Re changes the frequency response of the system. Several dB's is a rather big thing isn't it? The total Q of the system also changes. The point is: quantifying drivers or perceived quality of speakers without taking into account the amplifiers that drive them is like talking about a race car and it's abilities on the road and coming home to analyze it only talk about motor specs and forgetting all about the tires.
Damping factor
A clear example of how the amplifier and the speaker's impedance and Re changes the frequency response of the system. Several dB's is a rather big thing isn't it? The total Q of the system also changes. The point is: quantifying drivers or perceived quality of speakers without taking into account the amplifiers that drive them is like talking about a race car and it's abilities on the road and coming home to analyze it only talk about motor specs and forgetting all about the tires.
Damping ratio: Damping ratio - Wikipedia It's the inverse of Q (well, it differs by a factor of 2).
Ah I think I see the source of confusion. You are right that Qes and BL do not influence resonance frequency (though I never said that it does). I realize that I might have caused some confusion by mentioning 'reducing cone mass'. I forgot to mention: accompanied by an increase in compliance such that the resonance frequency is kept equal. When saying 'no bass' I meant due to the low system Q, with an equal resonance frequency.
Ah I think I see the source of confusion. You are right that Qes and BL do not influence resonance frequency (though I never said that it does). I realize that I might have caused some confusion by mentioning 'reducing cone mass'. I forgot to mention: accompanied by an increase in compliance such that the resonance frequency is kept equal.
When saying 'no bass' I meant due to the low system Q, with an equal resonance frequency.Well sure, because of (Rg being all impedances in series with Re):
Qes = ß ×( Re + Rg )/ BL²
In that case a distinction could be made between active vs passive speakers.
If I'm not mistaken almost everybody here listens to passive speakers, so it's good to point it out.
Btw. Good dynamics imho can only truly be generated by large surfaces (can be waveguide/horn/esl etc) accompanied by, sad but true, the bigger (built quality and size) woofers. I don't think it's electrical efficiency per se, but more acoustic efficiency that matters most.
For people that own relatively small speakers I feel it's paramount to try building a bigger and higher efficiency speaker. Think big, buy quality drivers.
When you figure out what drivers you need, KISS al the way up to the cross over parts, first shot 2nd order woofer, 1st order for the rest should suffice and hook it up to a good amp. There's quite some tweaking involved in choosing parts (better think cheap than exotic at first) but it's doable. From there on I'm pretty sure you'll not be concerned about dynamics anymore, or distortion. Most "distortionproblems" I encounter are caused by bad filtering, phase problems, just plain faulty drivers and non optimal placement in the room. It's been said before but if you're like me, or want to shortcut 25 years of trying differently, it's good to repeat;-)
Cheers
Qes = ß ×( Re + Rg )/ BL²
In that case a distinction could be made between active vs passive speakers.
If I'm not mistaken almost everybody here listens to passive speakers, so it's good to point it out.
Btw. Good dynamics imho can only truly be generated by large surfaces (can be waveguide/horn/esl etc) accompanied by, sad but true, the bigger (built quality and size) woofers. I don't think it's electrical efficiency per se, but more acoustic efficiency that matters most.
For people that own relatively small speakers I feel it's paramount to try building a bigger and higher efficiency speaker. Think big, buy quality drivers.
When you figure out what drivers you need, KISS al the way up to the cross over parts, first shot 2nd order woofer, 1st order for the rest should suffice and hook it up to a good amp. There's quite some tweaking involved in choosing parts (better think cheap than exotic at first) but it's doable. From there on I'm pretty sure you'll not be concerned about dynamics anymore, or distortion. Most "distortionproblems" I encounter are caused by bad filtering, phase problems, just plain faulty drivers and non optimal placement in the room. It's been said before but if you're like me, or want to shortcut 25 years of trying differently, it's good to repeat;-)
Cheers
Damping ratio: Damping ratio - Wikipedia It's the inverse of Q (well, it differs by a factor of 2).
Ah I think I see the source of confusion. You are right that Qes and BL do not influence resonance frequency (though I never said that it does). I realize that I might have caused some confusion by mentioning 'reducing cone mass'. I forgot to mention: accompanied by an increase in compliance such that the resonance frequency is kept equal.
Aha twice: I assumed with ratio you were using different lingo for the same thing (damping ratio = damping factor). Sorry for the confusion, my bad.
To take a really low impedance coil a step further it would need to be driven by a really low impedance source. Since active is the only viable way perhaps, this Q-/loss of bass issue could be dealt with easily.
Would a driver made this way by itself not have a lot less phase deviations etc?
Put differently: I'm guessing (bass) drivers (and perceived quality) are hardly going to get on a substantially higher level unless the driving force gets a from-scratch-redesign.
Writing it all down it seems it's filled with incompatibilities, so we might get stuck right here, true for sure.
Why am I so hung up on this you might ask?
Well, it's easy to notice that almost all drivers in a speaker can be attenuated, if needed, with a resistor or l-pad, without extreme penalties to the sound quality. Even though i don't like them at all, it's almost always better to have them there to adjust the level to the rest of the drivers. We all know. But when it concerns bass drivers we all know it's absolutely detrimental to the sound having even a small resistance in series. Under an ohm or so you're relatively oke, above and you're quickly hearing no bass, fluffy bass, take your favorite adjective for bad bass. No articulation, no drive, no dynamics. I'm sure that when you measure its frequency range it changed substantially, but not as bad as it sounds. It's still quite a decent speaker in a way. It is also why I mentioned Damping Factor and it's relatively useless means of examining expected bass quality.
Finally:
I modified 4 bass drivers from 8 to 2 and one to 0.5 ohm coil resistance. Only the coil was changed, no modification at any other physical quality was done. Mass might have changed a few grams max, but we're talking 100 grams or more mms (15" jbl and 15" rcf).
The perceived quality increase was the same as going from a few ohms in series with <your favorite woofer> to <your favorite woofer> without a few ohms. Great articulation, depth, thickness, sounds made "from the bottom up", if you catch my drift.
And they all do sound a lot more dynamic ("bigger") than they did.
Now I understand this might come over as a freak show to maybe some of you, with some pseudo science talk poured over it, maybe not. I certainly hope not. But I thought I should just add to this thread a significant factor which propably has not been tried yet to raise the quality, or lower the perceived distortion of drivers.
Since I'm not theoretically educated enough to explain this, I'm hoping someone here could shed some light on is?