An audio DSP chip from NXP Semiconductor NV (Eindhoven, The Netherlands) is able to boost the output power of small speakers used in smartphones and tablet computers by a factor of five while still preventing damage to the speakers.
The use of a CoolFlux DSP and optimization algorithms on the TFA9887 chip is able to drive 2.6 watts rms into micro speakers previously rated for 0.5 watts. This is claimed to give mobile phones, portable music players and tablets louder sound and deeper bass response. Safety features include adaptive excursion control and real-time temperature protection, and the TFA9887 monitors speaker behavior through current-sensing amplifier.
The TFA9887 optimizes the audio signal based on the movement of the speaker and because the speaker is protected it can more power across more frequencies to make the sound louder. The chip also provides clip avoidance to cope with sagging power supply and a dc-dc converter to maintain performance despite battery undervoltage.
The advanced, embedded algorithms require no separate licensing. Additional tools allow designers to customize audio sound quality and choose how to optimize mobile device performance.
Shawn Scarlett, director of marketing for mobile audio products talks about the TFA9887 audio IC
Motional feedback in loudspeakers has been "invented" many times. I "invented" such a system in 1973 and built a working prototype as a final year project for my Bachelor's degree. I even registered a provisional patent application at the UK Patent office. Shortly after, the Philips prduct was launched. A couple of years later I found out about the 1924 patent
That actually pre-dates the 1927 landmark "invention" of negative feedback by Harold Black
Actually, in a properly tuned 4th order vented box loudspeaker, at resonance the cone motion goes to zero and instead the output comes from the tuned port - box (Helmholtz) resonator.
But, tis is basically the only configuration where this occurs: Other configurations, such as infinite baffle and sealed box, will cause the voice coil motion to go to infinity.
Where this would be a Very Big Help is in hearing aids ;-)
Actually, motion control feedback goes back commercially to 1978 with the M&K "Volkswoofer" by Miller & Kreisel (I had one).
Prior to that, in 1969 my uncle, who was an RCA Engineer 1956-87, wrote his Master's thesis on a motion control feedback woofer system he designed, built and tested. (I still have a copy!) However, in his thesis he mentioned that for years, motion control had been used in the cutting head for record mastering, to reduce distortion.
Some 30+ years ago positional feedback (ie measure speaker element excursion) to linearize bass response was pionered by Philips. Our experiments to improve their rotten result simple endes up eventually bowling up the voice-coil. No DSPs available that could fix it at the time. However this shows that good ideas come back when both technology and market are aligned.
Chanj, you are correct that the physical size and magnetic power defines the limit of what the speaker can produce, however normal systems aren’t safe to operate at low frequencies. The size of the speaker and enclosure along with the behavior sound waves at the port will define a resonate frequency, at that frequency the speaker moves very easy – so a little bit of power can destroy the speaker. As a result micro-speaker systems have to filter out a wide range of frequencies, normally everything under 1 kHz, just to play it safe. The TFA9887 measures the resonate frequency and monitors the excursion, so you can put the bass back into the signal (no more 1 kHz high-pass filter). The result is that the speaker can produce all the sound it’s capable of.
The innovation here is that the actual speaker performance is used to adjust the volume to max at all times, but without overdriving the speaker coil. It also adapts to oddball situations, like for example something mechanically blocking the excursion of the speaker cone, which changes the actual impedance of the voice coil.
I think this technology has been used for decades by the TV ads. It was the only way they could rapidly get your attention by giving you that "boost" in delivered power so that you would take notice of the change.
Overall, a good clean approach to boost effective volume without adding distortion. It has been a long time coming to commercial products.
It seems like an interesting technology. I was taught that loudness is defined by the size of the speaker and the strength of the magnetic. In particular, the bass is generated by the moving of the speaker diaphragm. I will go check out more info from NXP.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.