Do electrostatic speakers/headphones really have no break-up modes?

[iqs2f1111]

From what I’ve seen, many websites and YouTube videos claim that electrostatic speakers/headphones are free from break-up because their diaphragms are driven uniformly across the entire surface, unlike cone speakers, which are driven only from a small central point.
SoundLab also asserts that electrostatic speakers have no “cone break-up distortion,” stating:

Low mass, distributed force: No other means of moving air has such inherent control. Distortion products are greatly reduced, even at high sound levels. As stated in non-scientific jargon, there is no “cone breakup distortion” since there is no cone to “break-up”.

However, according to this paper, the author measured the diaphragm’s motion using a laser system and found that break-up actually begins at around 375 Hz — well below the frequency where conventional cone speakers typically start to break up.
Vibration and sound radiation of an electrostatic speaker based on circular diaphragm

I wonder whether the idea that electrostatic speakers/headphones don’t have break-up is just a myth, or if the paper’s findings are inaccurate.

 

[Francis Vaughan]

OK, that is a really nice paper. Some seriously careful and non-trivial effort went into that.

If a diaphragm was somehow effectively massless, free at the edges, un-tensioned etc, an assertion of lack of break up might have more merit. But any real diaphragm isn't so. It has mass, is fixed at the edge, and in tension. So any idea that it uniformly moves is clearly wrong, and thus the logic that it should not suffer various planar vibration modes (aka breakup) is similarly fundamentally wrong. A circular membrane subjected to uniform force will curve exactly as the paper says. When the excitation matches the time time it takes a wave to propagate from the centre to the edge and back, or a harmonic thereof, it will breakup. Other modes will similarly become excited. Slight imperfections will add to the number of possible modes.
The paper, whilst demonstrating the value of their construction and measurement technique, doesn't measure any production electrostatic headphone. They built their own, and don't provide any information about membrane characteristics, tensioning or damping. So translating actual numbers to real life products should be done with care.

The paper does, not just in the measurements, but in the theory presented, show exactly why no real planar device is somehow intrinsically free of such breakup. There can be good reasons why, in practice, breakup is less of a problem, or intrinsic reasons why breakup can be better controlled or ameliorated in planar designs. But there is no magical freedom from them.
This is however not the point of the paper. They are describing how they have constructed and used what should be a valuable tool. The perforated Indium Tin Oxide electrode, allowing for optical measurement of the diaphragm. Something that could be applied to the design and measurement of new electrostatic headphones. They show how theory and their new device match. Theory includes these vibration modes. Nobody with any clue would ever think otherwise. The paper shows how they built a useful device that measures the reality, and that theory and measured reality match well. Thus is a vindication of their measurement technique.

 

[solderdude]

Electrostatic and planar magnetic (incl. isodynamic) do not move uniformly simply because the membranes are stretched and fixed near the edges.
So around the edges there is no movement (but in the case of electrostatic) there is still a drive signal there.
In planar magnetic (and isodynamic) the membrane is only driven where the magnets are. Here too the membrane does not move at the edges.
The membrane can move better/more towards the center. How far it can move depends on the membrane (it has to stretch a little).

Dan Clark has 'wrinkled' the membrane so it can move more uniformly.

Perhaps there is no break-up (as seen in membranes driven in a circle) but these drivers are plagued with lots of (poorly damped) resonances (the grass we see in the FR and ridges we see in CSD).
Also the HD800 ring driver is more uniformly driven but has other issues.

 

[DVDdoug]

SoundLab also asserts that electrostatic speakers have no “cone break-up distortion

Nothing is "perfect" in the real-physical-analog world.

The real issue is if the distortion is audible and overall performance. Laser measurements (etc.) might help with analysis, or during development to determine where distortion is coming from and it might be "interesting" but as consumers we only care about how the headphone sounds/performs.

Amir reviewed the Stax SR-0095 and it got "recommended only with EQ". He liked it after fixing-up the frequency response but he had to be careful with bass correction and volume to keep the distortion down.

There are good and bad headphones of all designs and in all price ranges.