Ideally damped speaker cone

[dennnic]

Hi,

Few weeks ago, I've connected my old Audiovectors to a tube amp for the first time and was left speechless.

Fast forward to a few days ago, when I found an interesting article (in the attachment below) regarding the mentioned subject. Article goes in great details about physics behind it.
Underdamped cone was showing considerably increased response towards a speaker resonant frequency. It was explained as a lack of an 'electronic brake', where the rigidity of a suspension overweighted effects of a voice coil. Those increased bass response was labeled as a distortion, a coloration. On the opposite side, overdamped cone showed lower SPL towards Fs, where a voice coil and magnets stopped the cone from moving more then it should.
Hence, a term of critical damping was mentioned as a solution, where driver won't oscillate due to it's suspension and will produce the needed SPL at the same time.

If we take a look from a perspective of a driver that perfectly follows the input signal, the less distortion should always be better. The only way to significantly decrease the effects from cone surround and spider, as I see it at least, could be to significantly increase the motor - overdamping the cone.

If we add a bit of parametric equalizer, do we get an ideal speaker, at the expense of an amplifier power? Overdamped cone should have a better transient response and eliminate effects of the suspension, leaving the driver more as a perfect piston. The part that doesn't fit in is why the cone tends to move less then needed with increasing the motor strength, and not follow the signal. Is a parametric EQ even a solution to an overdamped cone?

 

[Head_Unit]

Underdamped cone was showing considerably increased response towards a speaker resonant frequency. It was explained as a lack of an 'electronic brake', where the rigidity of a suspension overweighted effects of a voice coil... The part that doesn't fit in is why the cone tends to move less then needed with increasing the motor strength, and not follow the signal.

So in some classical electrical filter sense, we have
- An underdamped system moves quickly to the equilibrium point, but overshoots and oscillates about the equilibrium point thereafter.
- A critically damped system moves as quickly as possible toward equilibrium without overshoot.
- An overdamped system doesn't overshoot, but moves more slowly toward equilibrium than a critically damped system. Why? Eh, ya know, that's a good question. I've always thought of it as a huge magnet "choking off" the frequency response, which doesn't really explain WHY. More magnetic inertia or something? I should know this but never thought about it that way.*
However electric filters rising to equilibrium from static DC pulses are not the same as woofers producing bursts of low frequencies. In a different thread somewhere-I'll have to think where-someone posted waveforms showing that a somewhat underdamped Q was actually more faithfully in following a toneburst, because the overdamped case missed a lot of the initial rise. But but BUT I *believe* this effect will depend on the tone burst frequency versus the system resonance frequency.
(I'd LOVE to find a simulator where I could simulate woofer-in-a-box and then input a tone burst. I'd like to study the effects of over- and under-damped sealed and ported alignments on tone burst shape. Once upon a time you could do this laboriously via an inverse transform exporting from LEAP to LMS however those tools are rather dead now.)

@Overdamped cone should have a better transient response and eliminate effects of the suspension, leaving the driver more as a perfect piston.
Well, "perfect piston" sounds like a perfectly rigid cone, let's not talk about the piston. And suspension effects are not eliminated at all, they are still resisting motion. And as I note above, "better transient response" in an electric filter may not be better with music. Especially since you pay a HUGE sealed box size penalty go from from Qtc=0.85 or whatever to Qtc=0.5.

@ If we add a bit of parametric equalizer, do we get an ideal speaker, at the expense of an amplifier power?...Is a parametric EQ even a solution to an overdamped cone?
Well maybe kinda sorta. People would make a sealed box of Q=1.0 then apply EQ to get a total system response equivalent to Q=0.5. However Dick Small himself would be the first to point out that all those numbers are MODELS applicable only to SMALL-SIGNAL conditions. Under actual drive conditions those parameters are wiggling around some, while the electronic EQ is not. There will be some error happening, more and more at higher and higher sound pressures.


*While if you put a driver in a bigger sealed box the Q is lower and the frequency response "droops" more, that is using one identical woofer in different boxes. Putting a bigger magnet on the same cone/suspension does not really "cut the bass" by the way: yes, some simulations may appear to show that, but if you change the magnet and overlay the responses, what happens is the low frequencies stay at the same SPL but the curve rises up more and more into high frequencies. Referenced to the midband yes the bass appears to roll off sooner, but the midband level will be higher and more efficient, without losing bass efficiency.

 

[gene_stl]

Way back in the day there was an English made spikker that had sandwich cones. I think it was Leak. I listened to a two way someone was trying to sell me. It sounded pretty good for the early 70s. Had an interesting Formica finished cabinet too. Gave the overall impression of very high quality stuff. It had about a three inch cone tweeter which also was sandwich coned.

 

[dennnic]

...An overdamped system doesn't overshoot, but moves more slowly toward equilibrium than a critically damped system. Why? Eh, ya know, that's a good question. I've always thought of it as a huge magnet "choking off" the frequency response, which doesn't really explain WHY. More magnetic inertia or something? I should know this but never thought about it that way.*

This is most unusual, I bet the same phenomenon tend to "choke" the frequency response towards lower frequencies. If an amplifier output impedance has such a notable effect on a driver damping, it must have something to do with an interaction of a voice coil and magnets. Maybe a higher amplifier's damping factor induce higher back EMF and it's interaction with magnets on the return path. Just a wild guess, unfortunatelly.

There is a interesting information and graphs with transient response on quarter wave - open baffle theory part 1. It covers three Eminence drivers with Qts of 1.2, 0.5 and 0.3. They all overshoot, but the overdamped has the shortest settling time.

@Overdamped cone should have a better transient response and eliminate effects of the suspension, leaving the driver more as a perfect piston.
Well, "perfect piston" sounds like a perfectly rigid cone, let's not talk about the piston. And suspension effects are not eliminated at all, they are still resisting motion. And as I note above, "better transient response" in an electric filter may not be better with music. Especially since you pay a HUGE sealed box size penalty go from from Qtc=0.85 or whatever to Qtc=0.5.

I agree, piston would be the wrong term. The idea was to find a driver that would minimize the effect of a suspension and weight, by making those parameters proportionally smaller compared to a motor strength. If you add transient response, you get an interesting package. But, if that were true, I guess everyone would be making overdamped system with a simple DSP somewhere in the chain. Maybe it's just tough to sell that.

 

[Shefffield]

I recently learned that Thiele Small parameters should be taken with a large grain of salt. Had the suspicion for quite a while anyway.

I know, the clips are pretty long, but for me they had quite a lot of useful information. Even if his mindset is clearly pro audio, most of it applies to hifi just the same.

And, yes: Pros obviously talk about pistons all the time when they refer to cone-VC assemblies.


In short: an overdamped system seems to come to rest quicker, yes. I'm aiming for the same in my designs. But TSP don't tell much about system Q outside of a very specific (and small) voltage/power range. And check the "drivers are not flat" video to see what happens in cone, spider and surround when things get moving - especially with low Mms.

 

[dennnic]

I recently learned that Thiele Small parameters should be taken with a large grain of salt. Had the suspicion for quite a while anyway.

I know, the clips are pretty long, but for me they had quite a lot of useful information. Even if his mindset is clearly pro audio, most of it applies to hifi just the same.

And, yes: Pros obviously talk about pistons all the time when they refer to cone-VC assemblies.


In short: an overdamped system seems to come to rest quicker, yes. I'm aiming for the same in my designs. But TSP don't tell much about system Q outside of a very specific (and small) voltage/power range. And check the "drivers are not flat" video to see what happens in cone, spider and surround when things get moving - especially with low Mms.

I knew I should pick drivers by the color.

I've watched the transducers are not flat video and it brings more stuff to think about. Drivers I bought have a high magnet strength, but the light cone, at the same time. They are overdamped, where the Qts varies between 0.25 and 0.35. I can't really find a driver (apart for maybe those ment for a subwoofer usage) that has both a heavier cone and significantly oversized magnets. It begs the question why. Either there is something else that degrades the performance with those drivers, or there isn't too many people who would use a DSP, and hence pay for such a driver.

I guess the reason Headunit woudn't refer to drivers as pistons is that, at least for the common materials out there, they exhibit quite a non linear behaviour. Nevertheless, if one is looking for a most piston like driver out there, is oversizing the magnets and somewhat the cone weight the way to go? Thus, overdamping the cone and allow it to get the quickest transient response, in hopes to track the input signal as close as possible.
Provided one is comfortable with using DSP. It seems as if most drivers on the market are actually made for a passive systems, to be more cost effective at the same time.

 

[Jim Matthews]

To the OP : do you find the Audiovector design responding to amplifier choice? This may indicate other parameters at play.

I'm a fan of two way designs that leave the LF driver effectively "full range" with few crossover components.

Most of my favorite vacuum tube amps had falling HF response, which may complement an AMT HF unit.

 

[Pegwill]

Hi

I’ not sure if this has any relevance to the discussion but there was a KEF bass unit B139 that had I think a polystyrene piece as part of the cone. It was an oval speaker . It was used in the KEF Concerto model. I made a diy pair, back in the day as they say. Always liked the sound. You still might be able to find details if the information is of any use.

Regards


Regards

 

[Shefffield]

I knew I should pick drivers by the color.

I've watched the transducers are not flat video and it brings more stuff to think about. Drivers I bought have a high magnet strength, but the light cone, at the same time. They are overdamped, where the Qts varies between 0.25 and 0.35. I can't really find a driver (apart for maybe those ment for a subwoofer usage) that has both a heavier cone and significantly oversized magnets. It begs the question why. Either there is something else that degrades the performance with those drivers, or there isn't too many people who would use a DSP, and hence pay for such a driver.

I guess the reason Headunit woudn't refer to drivers as pistons is that, at least for the common materials out there, they exhibit quite a non linear behaviour. Nevertheless, if one is looking for a most piston like driver out there, is oversizing the magnets and somewhat the cone weight the way to go? Thus, overdamping the cone and allow it to get the quickest transient response, in hopes to track the input signal as close as possible.
Provided one is comfortable with using DSP. It seems as if most drivers on the market are actually made for a passive systems, to be more cost effective at the same time.

Hey,

same conclusion here. Forget the wannabe "data sheets", go by colour and feel. :-/

I do hope that my driver behave in a linear fashion most of the way. Nonlinearities = distortion ('Klirr' in German, hard to translate). Distortion can be measured relatively easily. And is, of course, heavily dependent on excursion (and temperature?).

Hence, use many smallish woofers that don't have to do much excursion for the desired SPL. I hope that such an approach allows me to avoid the most expensive drivers and still get good performance. Enough cone area combined with many motors also makes heavy EQing at the low end less demanding.

I might have missed if we talk about vented or closed subs. I only consider closed ones.


What is Q(uality)?

To my understanding, it describes the curve of the resonance we are looking at. Low Q describes a broad peak with low amplitude, while high Q describes a narrow peak with high amplitude. What do we want for a subwoofer?

Well, I guess that depends.

A low Q describes a well damped speaker. In our case electrically damped. If we mount a driver in a very small volume, Qtb will be high - and also at a high frequency. Higher frequency resonances carry less energy. My analogy is that the driver in the small cabinet has to work against a much stiffer 'air spring'. Maybe less electrical damping, but more mechanical damping.

The beauty of DSP is that we can easily roll over the high Qtb and equalize the frequency response to our desire. All we need is an amplifier that can put out enough volts and a driver that can take the pressure.


I recently wondered if such a setup with the cabinet/driver resonance in the middle of the XO transmission window should be avoided, as per the old school approach with all passive circuitry and no EQ. Given the experience with Zoé (see my avatar, digital speaker with massive 15" in ca. 30 l) it's no problem acoustically.

 

[dennnic]

To the OP : do you find the Audiovector design responding to amplifier choice? This may indicate other parameters at play.

I'm a fan of two way designs that leave the LF driver effectively "full range" with few crossover components.

Most of my favorite vacuum tube amps had falling HF response, which may complement an AMT HF unit.

Yes, there is a significant difference when tube amplifier is connected to vectors. Bass is significantly more present, it even sounds as it extends lower. On the other hand, when I switch back to solid state, timing seems better, there is bass and punch, but it doesn't sound as full as with the tubes.
There might be more to it, I'm just not aware as of now. I did try another tube amp before, and even a tube preamp, but the latest 300b tube amp got the sound just right.

I'm just finishing the latest project, that's also two way, but completely oposite, small full range that covers most of the spectrum and two huge woofers to support the LF.
Have you ever tried boosting the high frequency response of tubes? I'm considering boosting a few dB the falling response (over 7-8k Hz) of a full range driver.

Hi

I’ not sure if this has any relevance to the discussion but there was a KEF bass unit B139 that had I think a polystyrene piece as part of the cone. It was an oval speaker . It was used in the KEF Concerto model. I made a diy pair, back in the day as they say. Always liked the sound. You still might be able to find details if the information is of any use.

Way back in the day there was an English made spikker that had sandwich cones. I think it was Leak. I listened to a two way someone was trying to sell me. It sounded pretty good for the early 70s. Had an interesting Formica finished cabinet too. Gave the overall impression of very high quality stuff. It had about a three inch cone tweeter which also was sandwich coned.

Haven't seen those, especially for sale around here. Audiotechnology uses plastic cones, they claim it has a lower Q. Sandwitching layers could also add to the strength of a cone. I've even seen some ceramic cones for sales, but at gigantic prices. Haven't had a chance to listen to any of them though, except for plastic, paper and some mixes with kevlar.

 

[Pegwill]

Hi

A link to the KEF B139 is below

https://us.kef.com/pub/media/pdf/Datasheet_B139B_SP1044.pdf

While they are quite old you might be able to buy a pair at a reasonable price from the bay. Not sure if there would be any new ones around but Falcon Audio in the uk May have something similar.

Regards

 

[dennnic]

Hey,

same conclusion here. Forget the wannabe "data sheets", go by colour and feel. :-/

I do hope that my driver behave in a linear fashion most of the way. Nonlinearities = distortion ('Klirr' in German, hard to translate). Distortion can be measured relatively easily. And is, of course, heavily dependent on excursion (and temperature?).

Hence, use many smallish woofers that don't have to do much excursion for the desired SPL. I hope that such an approach allows me to avoid the most expensive drivers and still get good performance. Enough cone area combined with many motors also makes heavy EQing at the low end less demanding.

I might have missed if we talk about vented or closed subs. I only consider closed ones.


What is Q(uality)?

To my understanding, it describes the curve of the resonance we are looking at. Low Q describes a broad peak with low amplitude, while high Q describes a narrow peak with high amplitude. What do we want for a subwoofer?

Well, I guess that depends.

A low Q describes a well damped speaker. In our case electrically damped. If we mount a driver in a very small volume, Qtb will be high - and also at a high frequency. Higher frequency resonances carry less energy. My analogy is that the driver in the small cabinet has to work against a much stiffer 'air spring'. Maybe less electrical damping, but more mechanical damping.

The beauty of DSP is that we can easily roll over the high Qtb and equalize the frequency response to our desire. All we need is an amplifier that can put out enough volts and a driver that can take the pressure.


I recently wondered if such a setup with the cabinet/driver resonance in the middle of the XO transmission window should be avoided, as per the old school approach with all passive circuitry and no EQ. Given the experience with Zoé (see my avatar, digital speaker with massive 15" in ca. 30 l) it's no problem acoustically.

I remember seing an interview with Siegfrid Liknwitz, where he claims that the use of multiple woofers for the same frequency produce higher SPL levels, but at the expense of vertical sound dispersion. He claimed the existance of multiple lobes when using multiple drivers. But they would have needed to move less and settle faster.

Well, the idea was to discuss woofers in general, whether an overdamped woofer would always be the better woofer. I wasn't trying to put more variables (cabinet effects) in, since it could be quite different for closed, vented box or an open baffle.

By the way, very nice speaker you've got there. Is that a supertweeter on top?

 

[Shefffield]

I strongly believe that overdamped woofers are preferrable! That's how I understand the remarks about electrical control in the 'woofers are not flat' video. Maybe the 'over' is misleading and we should just call them 'tightly controlled'?

I had the pleasure to listen to a few feedback controlled speakers (Silbersand and restored/modified old Backes & Müller), and they are pretty much the best speakers I have listened to, period. How much 'dampening' does an active control via feedback loop provide?

Re Linkwitz, I only thought about subwoofers here. No side lobes to worry about here.

Thank you! That's a trophy from a hobby competition. Not active, and not even part of the system (we had to give it back so it could move on to the following year winner).

 

[dennnic]

I strongly believe that overdamped woofers are preferrable! That's how I understand the remarks about electrical control in the 'woofers are not flat' video. Maybe the 'over' is misleading and we should just call them 'tightly controlled'?

I had the pleasure to listen to a few feedback controlled speakers (Silbersand and restored/modified old Backes & Müller), and they are pretty much the best speakers I have listened to, period. How much 'dampening' does an active control via feedback loop provide?

Re Linkwitz, I only thought about subwoofers here. No side lobes to worry about here.

Thank you! That's a trophy from a hobby competition. Not active, and not even part of the system (we had to give it back so it could move on to the following year winner).

That's how the video sounded to me, too. However, would you then only use amplifiers with very high damping factor?

Can you explain feedback controlled speakers? I haven't heard of that concept.

I like the white cabinet, especially the part around midrange driver. Nice idea for a first price on a competition.

 

[Shefffield]

That's how the video sounded to me, too. However, would you then only use amplifiers with very high damping factor?

Can you explain feedback controlled speakers? I haven't heard of that concept.

I like the white cabinet, especially the part around midrange driver. Nice idea for a first price on a competition.

Yes, in the interest of high fidelity (exact reproduction) of the input signal I only consider amps with low output impedance appropriate.

Drivers that are loopback controlled have hardware built in that detects the actual cone movement. This can be a capacitive device (as has been used for dome tweeters) or a coil (standard for cone drivers). The detected movement is fed back to the amplifier to make sure that the cone moves exactly how the input signal demands. Any unwanted cone/diaphragm movement is actively suppressed by the amp.

The price was actually for sonic performance, not for looks.

 

[ZolaIII]

It's all cone material related and you ideally want impulse response to be exactly the same to original. Same high and time. But that still whosent achieved. Of course less rigid cones will do better regarding timing while more rigid with higher response... Only old paper cones do the timing right but they lack regarding how high it should be.
Take a look at the end of this video it's MDP Mica injected vs hand paper one and pretty much self explaining.

Best regards.

 

[JRS]

Hmmm. Not an EE, but I suspect that what you would end up with a more complex system, namely a sixth order system: two for the woofer itself, 2 for the box, and two (at a minimum) for the EQ via DSP. While you could likely shape the FR to whatever target (within the constraints of the drivers linear range and the power available, the transient response may actually suffer. I believe that WINISD could model this situation but you would have to estimate the output impedance of the amp at bass frequencies. It allows you to add s 2nd order EQ filter with several choices including Linkwitz transform. There are alignments such as quasi-Butterworths that help with the ringing, but the price you pay is a thirty six dB/octave roll off. Unfortunately I don't see a situation where you could improve the overall transient response, but the first to admit that my knowledge is limited.

 

[Jim Matthews]

Have you ever tried boosting the high frequency response of tubes? I'm considering boosting a few dB the falling response (over 7-8k Hz) of a full range driver.

I have, not with great success. My favorite tube amps had decent bandwidth, but power output fell at frequency extremes.

I used a high efficiency compression driver with a wideband 12" paper main driver. It did not produce a convincing stereo image *and* required separately powered subs.

The attempt to integrate these led me to MiniDSP and then to purely digital amplification where EQ is much easier.

 

[ZolaIII]

Out of curiosity I would try that with the largest one Foster has in current offering FE206En.

 

[JRS]

I strongly believe that overdamped woofers are preferrable! That's how I understand the remarks about electrical control in the 'woofers are not flat' video. Maybe the 'over' is misleading and we should just call them 'tightly controlled'?

I had the pleasure to listen to a few feedback controlled speakers (Silbersand and restored/modified old Backes & Müller), and they are pretty much the best speakers I have listened to, period. How much 'dampening' does an active control via feedback loop provide?

Re Linkwitz, I only thought about subwoofers here. No side lobes to worry about here.

Thank you! That's a trophy from a hobby competition. Not active, and not even part of the system (we had to give it back so it could move on to the following year winner).

Me too. I prefer nearly a 0.5 Q with infinite baffle alignment and an appropriately low F3, no cabinet to interfere, and no backwave cancellation like the open baffle systems. Tight, articulate and great reach because fs is fs for the woofer.