[Sandman65]

Do dynamic headphones have impedance phase angles or do they only present a simple resistive (vs. freq) load to amplification?

 

[Steve Eddy]

The former.
 
se

 

[Sandman65]

thx

 

[stv014]

Check the PDF files here for impedance and phase vs. frequency measurements of various headphones.
Here is a rough approximation of the HD598 with a simple circuit:

Note that dynamic headphone drivers are not only reactive, but also non-linear (i.e. the current they draw from a voltage source will have some THD and IMD products). Additionally, they can behave as a microphone, even if not as a very efficient one (dynamic headphone drivers and dynamic microphones are similar devices, just optimized for different applications). Finally, the characteristics of the drivers will change depending on whether the headphone is worn (which increases damping) or not. Even "burn-in" has some small effect on the impedance vs. frequency curve.

 

[OJNeg]

Hey stv, could you explain the circuit schematic in a bit of detail? I imagine that is an equivalent model of what the amplifier will see as the load, correct? How is it derived in this case?

 

[stv014]

As I noted, it is a very simple model, R1 and L1 simulate the resistance and inductance of the voice coil, while R2, L2, and C1 create the resonant peak at the primary resonance frequency of the driver (~90 Hz in this case). You can read an article on this model here. It is obviously not complete, and does not take into account the resonances of the enclosure, for example, nor any non-linearities. For simulating a load to an amplifier, the capacitance of the cable (typically at least a few hundred pF) is also relevant, as it may cause stability problems with a not sufficiently well implemented amplifier.
 
Quote:

How is it derived in this case?

 
Are you interested in how the component values can be calculated ?

 

[Chris J]

Quote:

Do dynamic headphones have impedance phase angles or do they only present a simple resistive (vs. freq) load to amplification?

 
Yes to the first part,
No to the second part.
 
Really just

 

[OJNeg]

Quote:

As I noted, it is a very simple model, R1 and L1 simulate the resistance and inductance of the voice coil, while R2, L2, and C1 create the resonant peak at the primary resonance frequency of the driver (~90 Hz in this case). You can read an article on this model here. It is obviously not complete, and does not take into account the resonances of the enclosure, for example, nor any non-linearities. For simulating a load to an amplifier, the capacitance of the cable (typically at least a few hundred pF) is also relevant, as it may cause stability problems with a not sufficiently well implemented amplifier.
 
 
Are you interested in how the component values can be calculated ?

 

Great article. I understand the graphs a lot better now.

 

[Sandman65]

Imaginary Numbers...yeah, I think I know what your driving at.
Once again I had a OCD moment where I thought if AMP output (SOA?) charts and headphone loads are known one could determine better system matching. 
(mostly to avoid clipping at typical playback volumes).
 
Seems more like Impractical or Intractable (numbers) vs. Imaginary?
 
I've moved past this latest execise in futility...

 

[Chris J]

Quote:

Imaginary Numbers...yeah, I think I know what your driving at.
Once again I had a OCD moment where I thought if AMP output (SOA?) charts and headphone loads are known one could determine better system matching. 
(mostly to avoid clipping at typical playback volumes).
 
Seems more like Impractical or Intractable (numbers) vs. Imaginary?
 
I've moved past this latest execise in futility...

 
LOL!
The imaginary number is just a very bad Engineering joke!
And everyone on Head Fi is OCD!
 
Impedance can be expressed mathematically as either:
67 Ohms, -15 degrees @ 2.7 kHz (for example)  can be read off an impedance plot
OR 
45 + j34 Ohms @ 3.4 kHz............j34 would be an imaginary number, 45 would be a real number.
 
It was a really bad joke!

 
But I see what you mean.....good luck getting SOA charts from an amp manfacturer!

 

[Sandman65]

I'll chalk it up to you having a "complex" sense of humor

 
tiny amps...
tiny speakers you wear on your head
 
tiny cables...
tiny white lies to steal your bread

 

[jcx]

do people really not know about HeadRoom's graphs, or that Tyll has continued measuring, contributing at http://www.innerfidelity.com/
 
http://www.headphone.com/learning-center/build-a-graph.php?graphID%5B0%5D=933&graphID%5B1%5D=703&graphID%5B2%5D=573&graphID%5B3%5D=2151&graphType=7&buttonSelection=Compare+Headphones
 

 
 
the graph doesn't range the axis but It looks like you can see a couple of bumps in the Ultrasone
 
the Senn HD600 has a nearly 2x impedance bump at the diaphragm mass-spring resonance frequency, maybe a breakup mode resonance bump and the generally rising impedance to the the right is likely voice coil inductance

 

[Chris J]

Quote:

do people really not know about HeadRoom's graphs, or that Tyll has continued measuring, contributing at http://www.innerfidelity.com/
 
http://www.headphone.com/learning-center/build-a-graph.php?graphID%5B0%5D=933&graphID%5B1%5D=703&graphID%5B2%5D=573&graphID%5B3%5D=2151&graphType=7&buttonSelection=Compare+Headphones
 

 
 
the graph doesn't range the axis but It looks like you can see a couple of bumps in the Ultrasone
 
the Senn HD600 has a nearly 2x impedance bump at the diaphragm mass-spring resonance frequency, maybe a breakup mode resonance bump and the generally rising impedance to the the right is likely voice coil inductance

 
The graph you posted does not show impedance phase angle, it shows impedance amplitude only.

 

[jcx]

single driver headphone terminal impedance is essentially  "minimum phase" - if you know the |z| vs frequency the phase is completely determined  - look up Bode's integral of phase: http://www.alcatel-lucent.com/bstj/vol19-1940/articles/bstj19-3-421.pdf
 
for a quick approximation the phase is a function of the |z| vs frequency slope - falling impedance with increasing frequency with 1:1 slope is "capacitive" has 90 degree phase "lag", conversely rising impedance with increasing frequency with 1:1 slope is "inductive" has 90 degree phase "lead"
 
you can verify "by eyeball" these results of the relations for the model circuit, magnitude, phase plots in post #4
 
 
multi-driver systems with crossovers can have more complicated "non-minimum phase" with "excess phase", diaphragm bending modes, breakup resonances may have some non-minimum phase component
 
the ideal would be to have the phase plot too, to see if there is much non-minimum phase component - but the graphs are still informative for many systems without the phase plot
 
the impulse response - even low enough frequency square wave response do contain the magnitude and phase information at the higher frequencies where the headphone transfer function, coupling to ear(microphone in head simulator) is more complicated but little of these details show on the electrical terminals of the headphone

 

[Chris J]

One thing the Inner Fidelity graphs are missing are impedance at frequencies above 20 kHz, in theory, a high enough phase angle at high frequencies could cause an amp to osciallate.
OTOH, the good amp designers have taken this "unknown" into account.