Solid State driver IEM on the horizon

[Dunring]

Just came across this and it looks pretty interesting. Using solid state drivers instead of the traditional designs. Going to set a Google alert in case something comes to market using this design. For IEMs/Ear buds they might use less power as well, for better battery life.

Solid-state driver earbuds are real and they're going to change portable audio

Thank you xMEMS Labs and thank you Creative Technologies

www.techradar.com

So far just a rendering prototype, but Creative Labs might play fast and loose when it comes to specs on sound cards, but they'll probably bring it to market.

 

[mhardy6647]

Piezoelectric?

 

[DVDdoug]

We'll have to see how they actually perform.

MEMS microphones are standard in cell phones and they are "very good" and cheap but I've never heard of them being used for pro recording.

If they technology takes-off they probably won't ALL be good, just as not all planar or electrostatic headphones are better than a good traditional headphone...

 

[oratory1990]

Piezoelectric?

based on piezoelectricity, yes.
But the material is PZT, not Quartz. It has very little in common with those piezo-buzzers.

 

[mhardy6647]

Sounds like it has everything in common with them.

 

[chi2]

"Sounds like" in the sense of "looks like" or in the sense of "actually does sound like"?

 

[oratory1990]

Sounds like it has everything in common with them.

other than the fact that they too rely on "voltage creates excursion", they really don't.
Common piezo speakers rely on deformation of a quartz crystal that is sliced as a disc from a grown monocrystal. Applying voltage to the quartz disc will cause it to bend.

MEMS speakers use lead-zirconate-titanate ("PZT") as an active layer. Applying voltage to PZT will cause it to contract.
By attaching the active layer to a passive layer (e.g. via a Sol-Gel process) this causes the passive layer to move in a pre-determined direction.
The active and passive layers are done in semiconductor processes (starting with silicon wafers), hence why they fall under the MEMS moniker.

xMEMS use these element to directly move air, meaning they need a certain area to produce a certain sound pressure.
Other manufacturers use the movement of these elements to move a more conventional diaphragm - this allows the active/passive layers to have a smaller area (and hence be produced significantly cheaper as you need less silicon per speaker) while retaining a large acoustically effective area.

 

[chi2]

@oratory1990 Thanks for the technical background. Are there any measurements available? And do you know if any of the implementations of this technology are capable of adequately reproducing frequencies in the 20-50 Hz range?

 

[oratory1990]

Are there any measurements available?

I'm not aware of any public measurements.
Samples are also very hard to obtain from xMEMS (and of course even harder for me, since I work for a direct competitor)

And do you know if any of the implementations of this technology are capable of adequately reproducing frequencies in the 20-50 Hz range?

On in-ear headphones, this depends mostly on how well the front volume is sealed - which is mostly determined by how well the eartip fits into the ear canal of the user (and how well the user manages to insert the eartip into their ear canal)
The passive layer on xMEMS is made from silicon and is rather stiff, leading to a relatively high acoustic impedance. Meaning the speaker imprints velocity on the sound field.
In other words: The movement of the speaker is determined by the voltage and the stiffness of the speaker, and is relatively unaffected by the acoustic load (e.g. the stiffness of the air in the front volume). This means that when the front volume is leaky (by an imperfect seal, or by controlled leakage through a venting hole) the movement of the speaker will be the same as before, and the sound pressure will go down (as it's easier for the air to pass through the leakage path as opposed to be compressed in the front volume).
We call this a "leakage intolerant" loudspeaker, as the SPL will drop with leakage.
If the stiffness of the loudspeaker is low enough then the movement of the speaker will (at least partially) be determined by the acoustic load. This means that when the front volume is leaky, the movement of the speaker will be higher (more excursion, since the stiffness of the air in the front volume is lower when the front volume is leaky), and it will produce a higher sound pressure at equal input. The sound pressure will still drop off (as air passes through the leakage path), but since there is higher sound pressure to begin with, the total loss of SPL will be lower.
We call this a "leakage tolerant" loudspeaker, as the SPL does not drop (or drops less) with leakage.

• MEMS loudspeakers that use silicon as the passive layer are leakage intolerant.
• BA speakers (typically using metal reeds / armatures) are typically leakage intolerant.
• Moving coil ("dynamic") speakers are typically more leakage tolerant, especially if they have a large diaphragm.
• Planar magnetic speakers of modern design are typically very leakage tolerant, especially if they have a large membrane. The 10mm planar speakers you find in e.g. a Tin P1 doesn't do better than an average dynamic speaker, for example. The 30mm speakers in the Audeze iSine however does excellent in that regard.

In short: No, the use of a silicon passive layer means they are not well suited for producing low frequencies in high-leakage scenarios.
I will say this though: You can make MEMS speakers with a passive layer that isn't silicon

 

[mhardy6647]

other than the fact that they too rely on "voltage creates excursion", they really don't.
Common piezo speakers rely on deformation of a quartz crystal that is sliced as a disc from a grown monocrystal. Applying voltage to the quartz disc will cause it to bend.

MEMS speakers use lead-zirconate-titanate ("PZT") as an active layer. Applying voltage to PZT will cause it to contract.
By attaching the active layer to a passive layer (e.g. via a Sol-Gel process) this causes the passive layer to move in a pre-determined direction.
The active and passive layers are done in semiconductor processes (starting with silicon wafers), hence why they fall under the MEMS moniker.

xMEMS use these element to directly move air, meaning they need a certain area to produce a certain sound pressure.
Other manufacturers use the movement of these elements to move a more conventional diaphragm - this allows the active/passive layers to have a smaller area (and hence be produced significantly cheaper as you need less silicon per speaker) while retaining a large acoustically effective area.

I think you might want to look a little harder at the old Motorola/CTS drivers, as you don't appear to know much about them.
Probably not your fault, though. Pretending that re-inventions of the wheel are "innovations" is practically a sport nowadays.

https://www.pispeakers.com/PIE_Speaker_Application_Note.pdf

You are correct, however, that many (perhaps all) use a diaphragm as a transducer (usually paper).

 

[oratory1990]

I think you might want to look a little harder at the old Motorola/CTS drivers, as you don't appear to know much about them.
Probably not your faul, though. Pretending that re-inventions of the wheel are "innovations" is practically a sport nowadays.

https://www.pispeakers.com/PIE_Speaker_Application_Note.pdf

You are correct, however, that many (perhaps all) use a diaphragm as a transducer (usually paper).

That's a free-field speaker though, not a microspeaker for occluded-ear applications

 

[mhardy6647]

I'm not sure if that's better or worse!

 

[oratory1990]

In a free-field speaker (especially for a tweeter) you don't need to worry too much about excursion below the resonance frequency. So the way the piezo actuators are attached to the structure that moves the diaphragm wouldn't be too complicated.
For occluded-ear applications, where SPL depends on excursion, that's of course a very relevant part of the design. The exact geometry of this attachment point becomes relevant. We have quite a lot of IP on this design aspect alone.