Electrostatic speakers as microphones?

[bolserst]

Originally Posted by bentoronto

This leads to something that has puzzled me for 50 years: where are the electret speakers?

An electret is a material that can retain a charge "permanently".
Basically the electrostatic equivalent of a permanent magnet.

So, an electret speaker or microphone can be thought of as an ESL that doesn't require a HV bias supply.
It's main use has been with microphones, but it has been used with some success in headphones.

[bentoronto]

Originally Posted by bolserst

An electret is a material that can retain a charge "permanently".
Basically the electrostatic equivalent of a permanent magnet.

So, an electret speaker or microphone can be thought of as an ESL that doesn't require a HV bias supply.
It's main use has been with microphones, but it has been used with some success in headphones.

Bolserst - thanks.

I think Stax made ESL headphones both ways.

If excellent for mics, why not speakers? Why not DIY?

Ben

[Marik]

Originally Posted by Bazukaz

Pressure gradient is difference in pressure, so how can we say "there is no pressure gradient" when there is sound in air??
When wavelength is small compared to glass size(which is true for all but the lowest frequency) different parts of glass will be hit at different times from typical sound sources so phase is not the same across the area.
I do not really get what do you mean by "strictly pressure operation" under these conditions ?

It would take me some time to describe it, so a quick google search shows this article, which explains it quite nicely (the first part):

Understanding & Using Directional Microphones

Originally Posted by Bazukaz

Also at 20 kHz 4-5 mm thick glass looks like a very hard wall with thickness equal to about a quarter the wave length(compare this to ~1.7meter@50Hz).

First, I am not sure we are talking the same thing. I am not trying to convince that such system would cover up to 20KHz (in this case microphone manufacturers would start making mics out of window glass). What I was saying is the glass window will have (in principle) flat response and have enough bandwidth for intelligibility of human speech.

Originally Posted by bolserst

Hello Marik,

Thanks for sharing your knowledge on ribbon microphones. I have no experience with them, but your description of the path length limitation makes sense. I have experimented with using small ESLs (1"x1" up to 3"x3") as microphones and the top end roll off was the same for all sizes if the diaphragm thickness and % open area of the stators was the same. Perhaps the reason for the difference in behavior is that the ribbon is mass controlled so the sound wave takes the path of least resistance and moves/diffracts around the ribbon-magnet structure and develops the delta pressure to move the ribbon. The ESL diaphragm doesn't present an obstruction to the sound wave like a mass controlled ribbon so moves with the air motion rather than reacting to a pressure difference due to the sound wave going around it. Thoughts?

As far as the mass controlled system concerned, even if you tune the system in the middle of the bandwidth (majority of the condenser mic capsules), in any case anything above that frequency is mass operated, and below, of course, stiffness operated. To make a flat response we will need to make the system resistance controlled (or EQ it).

It is hard to tell without seeing your ESLs, but I think in your case the stators work as resonators, which dominate the top response. Does it make sense?

Best, M

[Marik]

Originally Posted by bentoronto

I think Stax made ESL headphones both ways.

If excellent for mics, why not speakers? Why not DIY?

The spacing between diaphragm and stators is much smaller in headphones and especially condenser microphone capsules, so there is no need for a high polarization voltage. I don't think electret material could be charged to some 6KV required for ESL speakers.

Best, M

[bolserst]

Originally Posted by Marik

It is hard to tell without seeing your ESLs, but I think in your case the stators work as resonators, which dominate the top response. Does it make sense?

Hmmmm....I might be missing something, but I don't think this is the case. If you are familiar with equivalent circuit modeling, the appropriate acoustic model for the stators would be perforated plates which add some resistance and mass in series with the radiation impedance. I don't think there is any stiffness element in the model that would produce a resonance any where in the top couple octaves.

Unfortunately the microphone test panels I was playing around with are still boxed up somewhere due to a recent move. However, attached is a pic to give you an idea of what the stators look like. Large holes in thin metal to minimize added acoustic mass.

If interested, I can post some measurements later this month when I locate them.

[Kindhornman]

bolserts,
If I remember correctly most perforated screens such as you are showing have about a 50% open to closed surface area. You would think that would be a lumped impedance in your calculations when doing your modeling? Is it a factor or is the impedance low enough to air at the frequencies we are talking about in a mic to leave that factor out?

[Marik]

Originally Posted by bolserst

If you are familiar with equivalent circuit modeling, the appropriate acoustic model for the stators would be perforated plates which add some resistance and mass in series with the radiation impedance. I don't think there is any stiffness element in the model that would produce a resonance any where in the top couple octaves.

We also have an equivalent inductance of the opening aperture (holes in the stators), and equivalent capacitance of the air trapped in the chamber formed between stators and diaphragm. The resulting resonant frequency=1/2pi x √LC

Best, M

[bolserst]

Originally Posted by Kindhornman

bolserts,
If I remember correctly most perforated screens such as you are showing have about a 50% open to closed surface area. You would think that would be a lumped impedance in your calculations when doing your modeling? Is it a factor or is the impedance low enough to air at the frequencies we are talking about in a mic to leave that factor out?

As best I could tell, the main lumped impedance contribution from the stators is a small inductor(ie mass).
Once you start using diaphragms thinner than 1/4mil(6uM) it starts to to dominate the roll-off behavior.

Originally Posted by Marik

We also have an equivalent inductance of the opening aperture (holes in the stators), and equivalent capacitance of the air trapped in the chamber formed between stators and diaphragm. The resulting resonant frequency=1/2pi x √LC

Understood, the stator holes act like little slugs of mass(inductance) and their contribution can be quantified using peforated plate formulas in Beranek. As far as the idea of the stator and diaphragm forming a chamber that would have stiffness the slugs of air could resonate with in the top octave...I'm having a hard time thinking that is the case. I'd be more inclined to think any stiffness would just be included as part of the capacitor modeling the diaphragm stiffness due to tenstion. This resonates with the sum of the lumped masses(inductance) which include diaphragm mass, reactive part of radiation impedance, and the stator holes.

But, it is easy to test if there is a resonance in the top octave. We should see the phase wrap if it passes thru a resonance. I don't recall seeing this in any of my measurements, but I wasn't exactly looking for it either.

Question...if this was the case, would you expect to see a similar resonance in the top end of all ESL speakers? I'm 99% sure this is not the case for ESL speakers where the diaphragm is the driving force rather than the air motion thru the stators. I've been surprised before by the difference in behavior when a transducer is used as a microphone vs a speaker, so keeping an open mind...it's the only way to learn

[Bazukaz]

Originally Posted by Marik

We also have an equivalent inductance of the opening aperture (holes in the stators), and equivalent capacitance of the air trapped in the chamber formed between stators and diaphragm. The resulting resonant frequency=1/2pi x √LC

Best, M

I am not sure if this effect could occur in audio frequencies, but remember some people have reporting problems when stators are too thick compared to hole size. It is reported that stators should be thin enough(in the order of 1 mm) with holes ~>2mm or so. However I have never seen a proof of any kind that negative effects are caused by resonance of air inside cavity and not too high air resistance.
Basically theoretical prediction should be backed up by measurements of some sort availability of which I do not know.
I think modeling these effects theoretically would be insanely difficult as we have gas which behaves like fluid, so not only open area matters, but hole size, spacing between holes, its shape, sharpness of the edge etc etc.

Regards,
Lukas.

[Marik]

Originally Posted by bolserst

Understood, the stator holes act like little slugs of mass(inductance) and their contribution can be quantified using peforated plate formulas in Beranek. As far as the idea of the stator and diaphragm forming a chamber that would have stiffness the slugs of air could resonate with in the top octave...I'm having a hard time thinking that is the case. I'd be more inclined to think any stiffness would just be included as part of the capacitor modeling the diaphragm stiffness due to tenstion. This resonates with the sum of the lumped masses(inductance) which include diaphragm mass, reactive part of radiation impedance, and the stator holes.

Let's make it much simpler--forget all equivalents, just think of it as a cavity resonance. IOW, Hemholtz resonator... just a tiny one, tuned into HF.

Originally Posted by bolserst

But, it is easy to test if there is a resonance in the top octave. We should see the phase wrap if it passes thru a resonance. I don't recall seeing this in any of my measurements, but I wasn't exactly looking for it either.

The peak can also be out of band, but still affect things. All I can say, a pure and symmetrical fig8 system (i.e. single diaphragm with its back completely open and both sides having the same acoustical loading) of 1" and 3" wide (or diameter) cannot have the same top frequency response, unless something else is going on.

Originally Posted by bolserst

Question...if this was the case, would you expect to see a similar resonance in the top end of all ESL speakers? I'm 99% sure this is not the case for ESL speakers where the diaphragm is the driving force rather than the air motion thru the stators. I've been surprised before by the difference in behavior when a transducer is used as a microphone vs a speaker, so keeping an open mind...it's the only way to learn

Indeed, the same transducer as a speaker and as a mic is completely different. For example, say overall 5" wide true ribbon speaker will hardly produce anything below some 400Hz with some kind of authority. Our Samar MF65 ribbon microphone with about 3/4" wide motor has -0.4dB on 20Hz (!!!).

I won't be able to give you any analysis of what difference the stators make. I'd think they do, as any obstacle is... an obstacle, and if anything, has its damping properties--after all, this is customary to put damping circles in front of dome tweeters (look at the ScanSpeak D2904, or Seas 27TAFNC/G). Of course, the easiest and quickest way to know would've been just to measure with and without, but unfortunately, we cannot make the ESL to work without stators .

Originally Posted by Bazukaz

I am not sure if this effect could occur in audio frequencies, but remember some people have reporting problems when stators are too thick compared to hole size. It is reported that stators should be thin enough(in the order of 1 mm) with holes ~>2mm or so. However I have never seen a proof of any kind that negative effects are caused by resonance of air inside cavity and not too high air resistance.
Basically theoretical prediction should be backed up by measurements of some sort availability of which I do not know.
I think modeling these effects theoretically would be insanely difficult as we have gas which behaves like fluid, so not only open area matters, but hole size, spacing between holes, its shape, sharpness of the edge etc etc.

To put the things quite simply, this kind of resonator system is widely used in the microphone production and many commercial mics do use it for HF extension/correction. To start with, most of ribbon microphones do have it, including Coles 4038, RCA 44, RCA BK5, AEA, Melodium 42A, Oktava ML16, 19, 51, Reslo, etc. etc. etc.
Also, many condenser microphones capsules also have it, Oktava MC219, MC319, many Audio Technica's. Heck!--even venerable WM60/61 also do have it!

Unfortunately, for many reasons I won't be able to publish here development data of our own Samar MF65 ribbon mic to fully illustrate the process of frequency response tuning, but I can assure you, as Viagra salesmen use to say: "It works"

Best regards, M