Re: Payin my dues
Konnichiwa,
One criticsm, your solution assumes an overly resonant (eg. badly designed) loudspeaker load. It would be good if you could show the responses for a speaker(system) with Qm = 0.7....
Also, I'm not certain if your output variable is correctly chosen. I would suspect the current through the 6 ohm series resistor would be more apropriate....
Further, for interrest, which Qm are you simulating (I'm too lazy to work it out from the RLC equivalent).
Sayonara
Konnichiwa,
phase_accurate said:
One criticsm, your solution assumes an overly resonant (eg. badly designed) loudspeaker load. It would be good if you could show the responses for a speaker(system) with Qm = 0.7....
Also, I'm not certain if your output variable is correctly chosen. I would suspect the current through the 6 ohm series resistor would be more apropriate....
Further, for interrest, which Qm are you simulating (I'm too lazy to work it out from the RLC equivalent).
Sayonara
When we discussed this trick with the combined current/voltage drive for the first time we were talking about "ordinary" drivers and not ones with high mechanical damping.
In case of a driver that is sufficiently damped mechanically, a pure current-source would of course be used. The response would of course be the same as using a voltage-driven speaker, apart from the already discussed advantages.
The chosen parameters in my example:
Qm=10
Qe =0.6
fres=50 Hz.
Re=6 Ohm
Lvc=0.8 mH
Regards
Charles
Konnichiwa,
Okay, cool. What i want to avoid is for some ....... to come out of the woodwork and note "look at the step response for current drive", forgetting that in this case a requirement is to have a low Qm....
It might be interesting to try this for a Qm of around 1.7 - 2, as a number of suitable drivers are available for this....
Sayonara
phase_accurate said:
Okay, cool. What i want to avoid is for some ....... to come out of the woodwork and note "look at the step response for current drive", forgetting that in this case a requirement is to have a low Qm....
phase_accurate said:
It might be interesting to try this for a Qm of around 1.7 - 2, as a number of suitable drivers are available for this....
Sayonara
Another side benefit of the low-Qm/High Qe system:
When you need good mechanical damping at Fs, you're also freed up to design for less energy storage in higher bands.
Take cone edge termination, for example. If you try to design a voltage-source driver cone with proper transmission-line termination to prevent reflections, you generally end up with too much damping at Fs. But if your driver requires a low Qm, you can add measures of resistive termination at the cone edge.
The same holds true for spiders and surrounds. When they go resonant in a standard driver, a voltage-source amp can damp out the motion they transfer to the VC, but the fact is they're still flapping away out there, adding their spectra to the driver's summed output, causing midband ripple.
Again, if you've got latitude to mechanically damp a driver, you can make these extraneous radiators go away.
When you need good mechanical damping at Fs, you're also freed up to design for less energy storage in higher bands.
Take cone edge termination, for example. If you try to design a voltage-source driver cone with proper transmission-line termination to prevent reflections, you generally end up with too much damping at Fs. But if your driver requires a low Qm, you can add measures of resistive termination at the cone edge.
The same holds true for spiders and surrounds. When they go resonant in a standard driver, a voltage-source amp can damp out the motion they transfer to the VC, but the fact is they're still flapping away out there, adding their spectra to the driver's summed output, causing midband ripple.
Again, if you've got latitude to mechanically damp a driver, you can make these extraneous radiators go away.
Would any application of damped springs (i.e. coiled metal springs over dampers) be useful as mechanical damping in a driver? Cms could be varied on the fly if such a design was desired, and it would reduce oscillation more effectively than a spring-only based mechanical damping system - furthermore - IMO - it would sap less energy than a conductive coil former.
Konnichiwa,
I don't think we want more springs.
However, the Eckmiller driver incorporated originally "grease pot" type dampers, I'm sure we can use a modern polymer in a similar application. Use a traditional spider (as opposed to modern fabricated cloth - the older versions where called "spider" for good reasons) and use a suitable blob of polymer directly contanting each spider leg to apply added damping.
Sayonara
454Casull said:
I don't think we want more springs.
However, the Eckmiller driver incorporated originally "grease pot" type dampers, I'm sure we can use a modern polymer in a similar application. Use a traditional spider (as opposed to modern fabricated cloth - the older versions where called "spider" for good reasons) and use a suitable blob of polymer directly contanting each spider leg to apply added damping.
Sayonara
Sure, multiple springs in a driver is probably too complex and cost-inefficient. But is the spider not a spring? Suppose we replace the spider (which does 2 things - centers the voice coil and provides mechanical damping) with a coiled spring around the coil former, and some method to keep the former centered in the gap (many different ways to do it, some of them even noiseless! - Bill F. probably has a few ideas concerning this function).
Now we're in about the same situation, though now we have a Cms that is quite constant throughout an excursion range most drivers will never be able to reach. I suppose a metal spring is less self-damped than a spider, but we can damp its oscillation with a typical oil-filled damper, which can be noiseless (and definitely will be, if insulated).
Though I can see how this adds unnecessary complication to a speaker design...
Now we're in about the same situation, though now we have a Cms that is quite constant throughout an excursion range most drivers will never be able to reach. I suppose a metal spring is less self-damped than a spider, but we can damp its oscillation with a typical oil-filled damper, which can be noiseless (and definitely will be, if insulated).
Though I can see how this adds unnecessary complication to a speaker design...
J12 DISTORTION REDUCTION IN MOVING-COIL LOUDSPEAKER SYSTEMS USING CURRENT-DRIVE TECHNOLOGY, Mills, P.G.L., Hawksford, M.O.J., JAES, vol.37, no.3, pp.129-148, March 1989
DOWNLOAD
J14 TRANSCONDUCTANCE POWER AMPLIFIER SYSTEMS FOR CURRENT-DRIVEN LOUDSPEAKERS, Mills, P.G.L., Hawksford, M.O.J., JAES, vol.37, no.10, pp.809- 822, October 1989
DOWNLOAD
http://www.essex.ac.uk/ese/research/audio_lab/malcolms_publications.html#Journal
DOWNLOAD
J14 TRANSCONDUCTANCE POWER AMPLIFIER SYSTEMS FOR CURRENT-DRIVEN LOUDSPEAKERS, Mills, P.G.L., Hawksford, M.O.J., JAES, vol.37, no.10, pp.809- 822, October 1989
DOWNLOAD
http://www.essex.ac.uk/ese/research/audio_lab/malcolms_publications.html#Journal
Not to pile it on, but I continue to salivate about the possibilities that open up when you set out to design a premium driver from the ground up for dominant Qm.
The guiding philosophy is pure and righteous: acoustic/mechanical answers for acoustic/mechanical challenges.
Picking up where I left off, consider the advantages of using flow resistance damping directly behind the cone to create a desired Qt. Damping material could be selected that largely absorbs higher frequencies, greatly attenuating internal cabinet reflections and re-radiation through the cone. Place it close enough to the cone, and I imagine it would also speed the decay of standing waves on the cone itself.
As 454 hinted, there are also some interesting suspension ideas that are inherently too damping for voltage drive but might be just the thing for current drive. For example, you can eliminate a spider altogether and substitute ferrofluid in the gap, giving you radial centering, viscous damping/termination, zero self resonance/noise, and improved heat sinking in one package.
Obviously, since you needn't worry about Qe taxing the LF, gap flux can be very high. You can also enjoy the benefits of full-length faraday rings since eddy-current damping is no longer a limiting factor.
A few more minutes, and I could probably think of some more nice things to say. 🙂
I believe the state of the dynamic transducer art could definitely be advanced in this direction. (Or, some might say, history could be repeated.)
. . . And if someone were to bring such drivers to today's market, they would have the market all to themselves. 😎
The guiding philosophy is pure and righteous: acoustic/mechanical answers for acoustic/mechanical challenges.
Picking up where I left off, consider the advantages of using flow resistance damping directly behind the cone to create a desired Qt. Damping material could be selected that largely absorbs higher frequencies, greatly attenuating internal cabinet reflections and re-radiation through the cone. Place it close enough to the cone, and I imagine it would also speed the decay of standing waves on the cone itself.
As 454 hinted, there are also some interesting suspension ideas that are inherently too damping for voltage drive but might be just the thing for current drive. For example, you can eliminate a spider altogether and substitute ferrofluid in the gap, giving you radial centering, viscous damping/termination, zero self resonance/noise, and improved heat sinking in one package.
Obviously, since you needn't worry about Qe taxing the LF, gap flux can be very high. You can also enjoy the benefits of full-length faraday rings since eddy-current damping is no longer a limiting factor.
A few more minutes, and I could probably think of some more nice things to say. 🙂
I believe the state of the dynamic transducer art could definitely be advanced in this direction. (Or, some might say, history could be repeated.)
. . . And if someone were to bring such drivers to today's market, they would have the market all to themselves. 😎
And don't forget the usage of the (conducting !) coil-former as eddy-current brake !
Regards
Charles
Regards
Charles
But that saps quite a bit of power... I remember there was a man who built his own subwoofer with an unbroken brass voice coil former... I recall him saying that it required almost 5kW to get respectable output from it.phase_accurate said:
Then again, maybe he was just terrible at building drivers. 🙂
Konnichiwa,
Does it, I mean does it REALLY do that inherently?
UNBROKEN BRASS? And how about a small an weak magnet too? Seems such a VC former would be rather heavy.
I guess it's again not so much what you do, but how you do it....
Sayonara
454Casull said:
Does it, I mean does it REALLY do that inherently?
454Casull said:
UNBROKEN BRASS? And how about a small an weak magnet too? Seems such a VC former would be rather heavy.
I guess it's again not so much what you do, but how you do it....
Sayonara
Sorry, I should have said "non-split", as is the case with most conductive formers.
Well, please remedy my ignorance. 🙂
By all means.Bill F. said:
I think you have already given the answer by yourself, how it should be done: Make a slotted former, tap both sides of the slot and connect a resistor to it. Or use a material with desired conductance from the beginning and use a closed former !
Regards
Charles
Regards
Charles
Bill F. said:
I like the idea of the driver enjoying enough acoustic resistance
that the damping is just sub-critical.
Amen.
And not just behind the driver. Let's open our minds to the potential of flow resistant/acoustically absorbant materials in front of a wideband cone (around the perimeter), damping the fundamental resonance and absorbing the extraneous resonance signatures of the surround and standing waves along the outer circumference of the cone. I imagine this treatment could effectively treat beaming/lobing in the bargain, even as it eliminated diffraction effects from the cone/frame/baffle transition.
However, I can also imagine the audiophile outcry at the very thought of placing anything between any part of the cone and the listener...