[jcx]

it turns out that full amplitude 20 kHz Sine waves are seldom encountered in music - even if we'd prefer our amp's were able to drive our headphones to that extreme
 
estimates of music "power bandwidth" range from 3-5 kHz, even MC phono carts can't track at much higher slew rate - a slew rate limit of > 5 kHz Sine full scale Vswing is going to be near inaudible without special test waveforms

 

[kevin gilmore]

When i had the T2 on my bench (without a load) i measured it at over 150V/us.
I plan on measuring the amps i have with a 50pf load, and taking scope pictures.
But it is going to be a while.

 

[alcofribas]

This is a great thread for information.  Voltage, current, slew rate...  now what about what's at the other end ?  I've never seen an impedance curve for a Lambda or an Omega.  Surely such data must be available somewhere.  We know it's essentially capacitive, but are there surprise dips here and there ?    Say we choose Lambdas, does the shape of the curve vary a lot from model to model ?
 
 

 

[kevin gilmore]

Going to require special test gear to do that. Virtually 100% capactive which means that
the reactance varies linearly with frequency.  Starts at literally meghoms at DC, and goes
all the way down to under 18k ohms at 20khz. Moisture in the air has some impact on
this.

 

[alcofribas]

It can't be perfectly capacitive because there would be no transmission of energy.  But perhaps this imperfection shows up more as phase deviation than as magnitude deviation from the ideal 1/f.  After all the membrane should transmit energy rather evenly at all frequencies.
 
But it would be interesting to know what's going on for real.
 
 

 

[kevin gilmore]

You are forgetting about the space charge which is what moves the diaphram.
In the same way as the diaphram in an electrostatic microphone moves the
space charge to generate the output voltage.

 

[alcofribas]

I'm not sure what you mean.  I was just remarking that a purely capacitive load does not consume any energy, so it cannot give an entirely satisfactory electrical model for a device that moves the air around it. 
 
I don't know anything about what it takes to model electrostatic headphones (is there any literature on this on the Web?), but I'm sure the space charge is an important consideration, now that you mention it. 
 
My question is about the deviation of the phase+magnitude impedance curve from the capacitive ideal, whether this data is obtained theoretically (for a sufficiently sophisticated model) or through measurements.
 
 

 

[jcx]

transducer electrical to acoustic efficiency is at best low single digit %, often much less - its hard to "see" the acoustic radiation term in most audio transducers by looking at the electrical terminal impedance
 
(I think we can ignore high efficency horn/compression drivers in a headphone forum, also cavity resonance might just be visible as elecrtical impedance peaks in some iem)

 

[arnaud]

 
Quote:

transducer electrical to acoustic efficiency is at best low single digit %, often much less - its hard to "see" the acoustic radiation term in most audio transducers by looking at the electrical terminal impedance
 
(I think we can ignore high efficency horn/compression drivers in a headphone forum, also cavity resonance might just be visible as elecrtical impedance peaks in some iem)

I agree on radiation impedance discussion but can't we at least observe the fundamental mechanical resonance of the (acoustic loaded) tensioned diaphragm (around 70-80Hz in Stax headphones per check of Spl response on a dummy head)?
 
In the case of an electrodynamic tranducer, as you probably know, the electrical impedance measurement is used to estimates the free air resonance and mechanical damping of the driver. Admittidly, the impedance is flat and small otherwise at low frequencies so the mechanical resonance is clearly visible in the impedance curve (as well as acoustic resonances when the driver is acoustically loaded).
 
In the case of estat, this resonance will be swamped because of the huge capacitance then ?
 
 
 

 

[arnaud]

 
Quote:

I don't know anything about what it takes to model electrostatic headphones (is there any literature on this on the Web?), but I'm sure the space charge is an important consideration, now that you mention it. 

 
I was very curious about it as well and took a shot at it. It's far from good as I am missing a few bits of information and overly simplified. There was very little interest for it too so I kinda gave up on it...
 
Here's the summary video: 

 

[jcx]

yes you can see resonance impedance changes, but I claim these can be seen as lumped mechancial mass-spring models
 
mass-spring resonance can involve the box/cavity air spring as well as diaphram mass, suspenstion spring constant - but this is not really radiated aucoustic signal - most of the air spring energy stays local, is mostly restored to the mechanical resonance
 
even Helmholtz resonance is largley explained by the air mass, cavity spring constant # - the acoustic energy leaving the system as radiated sound is small for most transducers and hard to "see" at the electrical terminals
 
radiated acoustic energy may be too simplified a concept in the case of headphones where our ears are in the cavity, there may be multiple cavity modes as sound wavelength gets smaller than cavity dimensions

 

[arnaud]

Quote:

yes you can see resonance impedance changes, but I claim these can be seen as lumped mechancial mass-spring models

 
Ok, we're in agreement from the start here: indeed for simple mechanical / acoustic behavior like mechanical resonance of cone with spider / suspension system or even the first acoustic resonances (helmholtz e.g ported or sealed) can be viewed as mass-spring (single dof) systems and / or through electrical analogy. What you were implying (I though) is that any of these dynamics are not visible in the electrical impedance but they totally are for electro-dynamic. Maybe not in the case of ES if the conductance is soo high such as an increase of 40 ohm isn't visible if the baseline in the mega-ohm range as Kevin says...
 
 
Quote:

mass-spring resonance can involve the box/cavity air spring as well as diaphram mass, suspenstion spring constant - but this is not really radiated aucoustic signal - most of the air spring energy stays local, is mostly restored to the mechanical resonance
 
even Helmholtz resonance is largley explained by the air mass, cavity spring constant # - the acoustic energy leaving the system as radiated sound is small for most transducers and hard to "see" at the electrical terminals
 
radiated acoustic energy may be too simplified a concept in the case of headphones where our ears are in the cavity, there may be multiple cavity modes as sound wavelength gets smaller than cavity dimensions

Ok, I am in agreement too . Can't represent a radiation impedance as simple mass-spring model hence the electrical analogy is only valid at low frequencies. I can believe as well as the radiation impedance isn't visible from the electrical impedance graph if taken driving the transducer. But, I wonder if you could do something by reciprocity (e.g. measure the electrical current generated when the headphone is excited by an acoustic signal, e.g. using it as a microphone). I believe another senior headier was telling me about these methods in the case of large estat panels, I'd have to look it up.
 
In any case, I would really welcome someone with access to equipment to show some electrical impedance plot of an SR009 for example .

 

[kevin gilmore]

I did have a chance to look at my antique vector impedance meter, and the problem is that it tops out about 10kohms.
Which means i can't measure anything under about 10khz. What i need to find is one of the agilent digital things
which should be able to measure to at least 200kohms. Still won't do the low frequencies. The right answer is
both impedance and phase. Probably would have to build something custom to do it right. Maybe a clever
impedance converter tied to tyll's AP might work.

 

[sridhar3]

Great thread.  Subscribed.