So Audio Technica released new headphones that they claim not to use any D/A chip. So they just pass the digital audio to the voice coil directly.
More here: ATH-DSR9BT Wireless Over-Ear Headphones with Pure Digital Drive || Audio-Technica US
I looked up on the internet trying to find more details about this technology, but only found this video from Clarion https://www.youtube.com/watch?v=HyhubYtrQzA
Which is even more counfusing... no amplifier?
So, does anyone have more knowledge about this, or is it just typical snake oil?
More here: ATH-DSR9BT Wireless Over-Ear Headphones with Pure Digital Drive || Audio-Technica US
I looked up on the internet trying to find more details about this technology, but only found this video from Clarion https://www.youtube.com/watch?v=HyhubYtrQzA
Which is even more counfusing... no amplifier?
So, does anyone have more knowledge about this, or is it just typical snake oil?
> they claim not to use any D/A chip.
This appears to be true.
FWIW, AT&T Bell Labs did this around 1970 (or before), and more directly. The motor was electrostatic diaphragm. There were eight segments with areas of 1, 0.5, 0.25, 0.125... An 8-bit (telephony) parallel digital audio signal was applied, MSB to the largest segment, LSB on the tiny segment. The receiver cap summed and low-passed the segments. It worked, a fine telephone. But at the time it had no advantage, nothing worth miniaturizing a rack of gear to a chip.
> So they just pass the digital audio to the voice coil directly.
Not quite that simple. Instead of 8 (or 16 or 24) binary-sized segments, they use a 4-winding coil, and process the 16/24 bits down to four more complex drive signals. (Clarion is promoting a form with six windings.)
FWIW, all such schemes simply move "the analog problem" from analog in voltage to analog in time. If the time periods are off, that is equivalent to a DAC's nonlinearity. It also echos DAC-users' concern with (time-)jitter. Since with fast clocks and stable oscillators, we really can do good with time, this may be better.
It is also entirely switching on the electronic side, so should give the good part-volume battery efficiency of Class D.
> just typical snake oil?
You have to squeeze them snakes to lubricate sales. Do you care Garter Snake oil or Hognose Snake oil?
This appears to be true.
FWIW, AT&T Bell Labs did this around 1970 (or before), and more directly. The motor was electrostatic diaphragm. There were eight segments with areas of 1, 0.5, 0.25, 0.125... An 8-bit (telephony) parallel digital audio signal was applied, MSB to the largest segment, LSB on the tiny segment. The receiver cap summed and low-passed the segments. It worked, a fine telephone. But at the time it had no advantage, nothing worth miniaturizing a rack of gear to a chip.
> So they just pass the digital audio to the voice coil directly.
Not quite that simple. Instead of 8 (or 16 or 24) binary-sized segments, they use a 4-winding coil, and process the 16/24 bits down to four more complex drive signals. (Clarion is promoting a form with six windings.)
FWIW, all such schemes simply move "the analog problem" from analog in voltage to analog in time. If the time periods are off, that is equivalent to a DAC's nonlinearity. It also echos DAC-users' concern with (time-)jitter. Since with fast clocks and stable oscillators, we really can do good with time, this may be better.
It is also entirely switching on the electronic side, so should give the good part-volume battery efficiency of Class D.
> just typical snake oil?
You have to squeeze them snakes to lubricate sales. Do you care Garter Snake oil or Hognose Snake oil?
Some more info here:
https://www.trigence.com/dnote
https://www.trigence.com/headset
The chip type number is DN3012, DN3032, DN3052, controlled by a DU1213. The 3rd digit apparently refers to the supply voltage, 1.8V, 3.3V or 5V.
The bluetooth speaker also looks interesting:
- 8.8W + 8.8W from 5V or USB bus power*
- Multi coil speaker (8-coil, 6-coil, 4-coil, 3-coil, 2-coil)
- 24-bit 96kHz USB audio interface
*Instantaneous Maximum Output Power
https://www.trigence.com/dnote
https://www.trigence.com/headset
The chip type number is DN3012, DN3032, DN3052, controlled by a DU1213. The 3rd digit apparently refers to the supply voltage, 1.8V, 3.3V or 5V.
The bluetooth speaker also looks interesting:
- 8.8W + 8.8W from 5V or USB bus power*
- Multi coil speaker (8-coil, 6-coil, 4-coil, 3-coil, 2-coil)
- 24-bit 96kHz USB audio interface
*Instantaneous Maximum Output Power
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.