I am doing some volunteer stuff with ALMA International (ALMA INTERNATIONAL - Home) for the next AISE convention in Jan. As those of you familiar with ALMA know, this group has been mostly the domain of the manufacturers of raw drivers. The theme is "Revolution of the Signal Chain". That got me thinking about what the next revolution in speakers could be.
Drivers are essentially a "dumb" component. They accept a signal and react to it. They have no built in protection capabilities or compensation capabilities. However, there are speaker systems that provide self-compensation (usually by DSP) by including their own mics and/or built-in amps that try to measure coil heating. That got me thinking as to how this might change if the drivers themselves could be made into a "smart" component. In other words, what happens to the design of speaker systems when the driver becomes a component that can provide information to the system it is placed into.
These days, many measurement microphones have a built in chip with TEDS (Transducer Electronic Data Sheet). The TEDS data can provide sensitivity info, model info, calibration data, etc to the measurement system. Perhaps a driver or speaker system can contain similar information (e.g. TS parameters) that the amplifier can use to provide some compensation. I can't say what benefit this would provide off the top of my head. However, two thoughts come to mind.
1) Perhaps a box can accept a wider tolerance of driver variation with electronic compensation circuits.
2) Perhaps a speaker system can vary its internal geometry (similar to the way some line arrays do) to compensate for driver variation.
Obviously, there would be additional costs involved in using a smart speaker. Are there benefits worth exploring though?
Information such as TEDS data is static. In other words, it doesn't change until it is explicitly rewritten. What if MEMS sensors are use inside drivers to provide real-time feedback. There is a new I3C standard from the MIPI alliance that is meant for use with IOT sensors that might be handy here. You could have a sensor for heating. Perhaps another to measure excursion, etc.
As I said initially, this is just a "what if" exercise. However, if there are enough positive "what-if" outcomes, I might try to put a panel session together for the AISE conference in January.
OK, let the slings and arrows fly!
Drivers are essentially a "dumb" component. They accept a signal and react to it. They have no built in protection capabilities or compensation capabilities. However, there are speaker systems that provide self-compensation (usually by DSP) by including their own mics and/or built-in amps that try to measure coil heating. That got me thinking as to how this might change if the drivers themselves could be made into a "smart" component. In other words, what happens to the design of speaker systems when the driver becomes a component that can provide information to the system it is placed into.
These days, many measurement microphones have a built in chip with TEDS (Transducer Electronic Data Sheet). The TEDS data can provide sensitivity info, model info, calibration data, etc to the measurement system. Perhaps a driver or speaker system can contain similar information (e.g. TS parameters) that the amplifier can use to provide some compensation. I can't say what benefit this would provide off the top of my head. However, two thoughts come to mind.
1) Perhaps a box can accept a wider tolerance of driver variation with electronic compensation circuits.
2) Perhaps a speaker system can vary its internal geometry (similar to the way some line arrays do) to compensate for driver variation.
Obviously, there would be additional costs involved in using a smart speaker. Are there benefits worth exploring though?
Information such as TEDS data is static. In other words, it doesn't change until it is explicitly rewritten. What if MEMS sensors are use inside drivers to provide real-time feedback. There is a new I3C standard from the MIPI alliance that is meant for use with IOT sensors that might be handy here. You could have a sensor for heating. Perhaps another to measure excursion, etc.
As I said initially, this is just a "what if" exercise. However, if there are enough positive "what-if" outcomes, I might try to put a panel session together for the AISE conference in January.
OK, let the slings and arrows fly!
Thanks. Yes, some of these are coming back to me. Times were different. Separates ruled. IoT is really dropping the price and size of sensors and DSP is no longer cost prohibitive. These older designs seemed to mostly focus on using motional feedback to improve linearity of bass. I think there is more that can be done now. Also, if the driver market standardizes on the type of data that can be delivered and the format (e.g. I3C) it might be possible to extend such techniques to "smart" separates.
Rythmik Audio uses what they call "Direct Servo" technology which is a separate sensing coil on the voice coil former to get direct feedback and use that to compensate the drive signal.
Rythmik Audio • Servo subwoofer products
Theoretically, this could be applied to smaller drivers, although it is only indirectly sensing the displacement of the voice coil/cone and would obviously not be able to account for any cone breakup modes (which you conveniently don't really need to worry about with subwoofers).
Rythmik Audio • Servo subwoofer products
Theoretically, this could be applied to smaller drivers, although it is only indirectly sensing the displacement of the voice coil/cone and would obviously not be able to account for any cone breakup modes (which you conveniently don't really need to worry about with subwoofers).
Unless I am missing something, the Rythmik solution is just a new spin on the earlier solutions. It is interesting but not revolutionary. I am interested in discussing what revolutions might be possible with other types of sensors or data that is stored within the driver.
Imagine an amp that EQs its output based on the stored frequency response data in the driver. This could have several benefits. The speaker system can accept a wider range of variation in drivers. This may help with using second sources, field repairs, etc.
Perhaps a heat sensor in the motor structure can tell an amp when to limit its output. This could allow operation at the physical limits of the driver without fear of failure.
Maybe a sensor can detect rubs and buzzes to alert the user that a repair is needed.
I am curious to see what other scenarios people can come up with.
Imagine an amp that EQs its output based on the stored frequency response data in the driver. This could have several benefits. The speaker system can accept a wider range of variation in drivers. This may help with using second sources, field repairs, etc.
Perhaps a heat sensor in the motor structure can tell an amp when to limit its output. This could allow operation at the physical limits of the driver without fear of failure.
Maybe a sensor can detect rubs and buzzes to alert the user that a repair is needed.
I am curious to see what other scenarios people can come up with.
Not necessarily an 'active' idea, but a possibility I wonder about is applying computer heat dissipation tech to speakers.
Specifically, would this give better power handling to small voice coil / wide bandwidth drivers?
passive versions include:
https://www.quietpc.com/nof-icepipe
Specifically, would this give better power handling to small voice coil / wide bandwidth drivers?
passive versions include:
https://www.quietpc.com/nof-icepipe
You didn't miss anything, but while it may not be revolutionary they have apparently implemented it well.
A heat sensor would be quite practical, but not very revolutionary
Going back to basics - how does a loudspeaker work? The cone moves by modulating a current through a coil in a fixed magnetic field. Most assume the magnetic circuit is good enough and focus on other aspects of the driver, but what can be done to improve the magnetic circuit? Instead of a permanent magnet why not use a field coil or multiple field coils? The magnetic field could be changed/modulated (current controlled) in an attempt to minimize undesirable characteristics or produce a larger, more linear magnetic circuit. Think Maglev train propulsion system but for a voice coil!
To take it one step further, and borrowing from what hollowly said, the field coil and pole piece could be temperature regulated (water cooled) to obtain stable performance.
Implementing the feedback loop is a whole different story, but this is just a 'what if' idea.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.