[PhonoPhi]

IEMs that at the end of the day end up being worse at following a target curve than single-driver IEM (Etymotics). I do not see why multi-driver IEMs are a thing, when you still need to optimize how you use a single driver for full-range, then you add more and make your design less coherent and harder to optimize. To be honest, EQ ends up being one of the best if not the best improvements you can get for your system, be it headphones, speakers, or IEMs.

BA drivers are limited in their range by design.
Also several BA drivers working together give a similar effect as a violin section vs. a single violin.
The coherence is definitely a limitation, but with more overlaping driver range (with more drivers), it can be partially mitigated.

 

[PhonoPhi]

The impedance values at resonances are actually higher, because admittance on the electrical side shows as impedance on the mechanical side:

Zm = (Bℓ)² / Ze.​

Here the constant (Bℓ)² is just the squared force factor Bℓ of the transducer voice coil. Mechanical resonances happen at frequencies where the mechanical impedance Zm has its local minimum (maximum mechanical admittance Ym = 1/Zm meaning the mechanical parts can move freely without much energy losses). On the electrical side this shows as maximum electrical impedance Ze (mimimum electrical admittance Ye = 1/Ze).

Since mechanical resonancies can only increase the electrical impedance, the minimum electrical impedance of a transducer is higher than the voice coil resistance. At some frequencies far from the driver resonance the highest frequencies where the voice coil inductance starts to "kick in" the impedance can reach this minimum.

It is true that the impedance increase appreciably near the resonance (at the lower energy side), as with any resonators/oscillators.

Then at the resonance, the energy consumption is at maximum, and the impedance goes to lower values. The linear scale of the resistance, most commonly plotted, does not make it as apparent as the pre-resonance increases.

 

[71 dB]

It is true that the impedance increase appreciably near the resonance (at the lower energy side), as with any resonators/oscillators.

Then at the resonance, the energy consumption is at maximum, and the impedance goes to lower values. The linear scale of the resistance, most commonly plotted, does not make it as apparent as the pre-resonance increases.

I have hard time following your logic here. Seems you are talking about temporal effects of some sort.

 

[PhonoPhi]

I have hard time following your logic here. Seems you are talking about temporal effects of some sort.

Here is the graph superimposing BA resonances and impedance.

Note the shift, and if more energy could be supplied, the resistance at resonances could go to very low values.

 

[peskypesky]

IEMs that at the end of the day end up being worse at following a target curve than single-driver IEM (Etymotics). I do not see why multi-driver IEMs are a thing, when you still need to optimize how you use a single driver for full-range, then you add more and make your design less coherent and harder to optimize. To be honest, EQ ends up being one of the best if not the best improvements you can get for your system, be it headphones, speakers, or IEMs.

Yeah, I've gone back to single dynamic driver IEMs lately. I experimented with some hybrids, but the dd's sound as good or better and are less expensive. At least from the ones I've bought.

 

[castleofargh]

Here is the graph superimposing BA resonances and impedance.
Note the shift, and if more energy could be supplied, the resistance at resonances could go to very low values.

I has confusing.
What is that graph in black exactly?


Here is what I think I understand about resonance:
At electrical resonance, I don't even need to summon my inner Pythagoras to get the vector, as total impedance equals resitance. So of course that's where we get lowest impedance and max signal can pass(more power).
BUT! That's if we only consider the electrical circuit as a cause of impedance.

Mechanical resonance is when the signal passes the best too, but mechanically. For dynamic drivers, the coil moving in a magnetic field will generate electrical signal opposing the original signal. Typically that's where we get the biggest impedance bump in the audible range(at low freq). And that's what @71 dB is saying.

For BA drivers, it's different in many ways, even though the laws of physics probably still hold
First, the coil doesn't move. So it's probably not a design as good at sending back EMF(my uneducated guess).
Single BA drivers do tend to have a measured impedance graph shaped like the blue line(with lowest impedance in the bass). The impedance values are a different matter as I doubt I have ever measured a single BA IEM with such a massive deviation in the midrange. But the general shape is a good basis to think about BAs and measured impedance.

Not sure I’ve contributed anything, please go on.

 

[71 dB]

Here is the graph superimposing BA resonances and impedance.
Note the shift, and if more energy could be supplied, the resistance at resonances could go to very low values.

I have been talking about dynamic drivers. I don't know much about BA drivers because I don't use IEMs. Something about BA causes the impedance drop at the maximum SPL frequency, but I don't know why because I have never analysed BA's. Sorry.

 

[PhonoPhi]

I have been talking about dynamic drivers. I don't know much about BA drivers because I don't use IEMs. Something about BA causes the impedance drop at the maximum SPL frequency, but I don't know why because I have never analysed BA's. Sorry.

All-BA designs were discussed, in the context of Andro. They do have very pronounced resonances compare to DDs, where the resonances of a thin light membrane is likely beyond the audible range.

I has confusing.
What is that graph in black exactly?


Here is what I think I understand about resonance:
At electrical resonance, I don't even need to summon my inner Pythagoras to get the vector, as total impedance equals resitance. So of course that's where we get lowest impedance and max signal can pass(more power).
BUT! That's if we only consider the electrical circuit as a cause of impedance.

Mechanical resonance is when the signal passes the best too, but mechanically. For dynamic drivers, the coil moving in a magnetic field will generate electrical signal opposing the original signal. Typically that's where we get the biggest impedance bump in the audible range(at low freq). And that's what @71 dB is saying.

For BA drivers, it's different in many ways, even though the laws of physics probably still hold
First, the coil doesn't move. So it's probably not a design as good at sending back EMF(my uneducated guess).
Single BA drivers do tend to have a measured impedance graph shaped like the blue line(with lowest impedance in the bass). The impedance values are a different matter as I doubt I have ever measured a single BA IEM with such a massive deviation in the midrange. But the general shape is a good basis to think about BAs and measured impedance.

Not sure I’ve contributed anything, please go on.

What is shown in black is effectively the sound produced by the driver at frequency sweep.

Here is more detail on these graphs, and more details on the BA design:
https://mynewmicrophone.com/the-complete-guide-to-balanced-armature-iems-earphones/

Indeed, not the coil itself is moving, but pretty stiff armature, and the resonance frequency is in the audible range.
With the resonance limitations, it is hard to make a single all-range BA.

Lastly, the impedance rises before reaching the resonant frequency. At the resonant frequency, the energy consumption can be huge (that is why distortions discussed), so if more energy will be supplied - the drivers can disintegrate, while the impedance is the lowest.
So good BAs are a lot about dampening, make them more manageable and smoother. There is a fine line in the dampening before BA start to sound like DDs (often the case with Sonion ones).

 

[71 dB]

Clearly I have been ignorant thinking the theory of DDs can be applied to all headphones... ...obviously what I said does not apply to BAs.

 

[magicscreen]

Audiophile myth: Tube amplifiers have a warm sound only because the tubes have light.
But if I listening with closed eyes or opened eyes, there is no difference.
Debunked & demystified!

 

[bigshot]

Tube amps are all over the place. The best of them sound as clean as solid state amps.

 

[PhonoPhi]

Audiophile myth: Tube amplifiers have a warm sound only because the tubes have light.
But if I listening with closed eyes or opened eyes, there is no difference.
Debunked & demystified!

A good one (as a "demistification" pun)

Tube amps are all over the place. The best of them sound as clean as solid state amps.

What about this science: https://producerhive.com/ask-the-hive/odd-vs-even-harmonic-distortion/

Do best clarinets sound like oboes?

 

[PhonoPhi]

Clearly I have been ignorant thinking the theory of DDs can be applied to all headphones... ...obviously what I said does not apply to BAs.

Your point was very helpful - pre-resonance rise in impedance is important, and may dominate for well-dampened BAs. I think now I got an explanation, why some all-BA designs behave differently. Thank you!

 

[bigshot]

Not all tube amps have euphoric distortion. Some are clean. Tube amps are all over the place.

 

[PhonoPhi]

Not all tube amps have euphoric distortion. Some are clean. Tube amps are all over the place.

True, from a consumer point of view.
Yet, if to talk about the science - the physics of overtones is there for lamps. Then the overall designs can either supress/minimize it or take an advantage of it. The former is hard to understand, given the efficiency of solid-state electronics.