Headphone Impedances

[joblessalex]

I have a pair of headphones at 17 ohms. They are being driven by what I am assuming is audiofoolery grade equipment.... it clearly states 600 ohm amp. Amplifiers are measured in watts, not output impedance. The headphones work, and are indeed loud enough to vibrate my ears, which is way too loud. Shouldn't a 17 ohm load cause distortion due to drawing more power than the system puts out?

[ConKbot]

Especially on consumer sound equipment, a 600 ohm amp means it capable of driving 600 ohm earphones. Not that it was meant for them. 8-32 ohm earphones are the usual range for stuff so you need an high voltage swing on the output to get the same power into a 600 ohm load.
And provided the amp is stout enough, as you've discovered, it means it can drive lower impedance phones quite loudly.

[JVR]

Also remember that impedance != resistance.

[mariush]

The text on the box says "support for 192 ...  ,  a 600 ohm amplifier ..."

It doesn't necessarily mean the headphones HAVE a 600 ohm amplifier inside them, just that they "support" a 600 ohm amplifier.

[JackOfVA]

I measured the impedance (|Z| and phase angle) of a Sennheiser HD201S headphone today to see how much the impedance varies with frequency and how inductive it is.

Plot  below is taken with an HP 4192A LF analyzer, measurements every 20 Hz from 30 Hz to 20 KHz. The data is for one channel of the headphones, with  the phones resting on a wooden bench top, open as much as they will open. There will be some change in Z and theta between this test condition and wearing the headset, but I would not expect a drastic difference.

DC resistance measured at 27 ohms.

There's not much change in |Z| over the frequency range, and overall the headphone transducer is not all that inductive. The coil resistance dominates the impedance over this frequency range.

Also somewhat surprised not to see more pronounced resonance points.

I forgot to label the graph lines - black is |Z| (left Y axis) and red is phase angle (right Y axis)

[krivx]

Interesting! This just shows the electrical properties of the coil - I would imagine the acoustic output of the speaker probably has quite a few more resonances

[JackOfVA]

Interesting! This just shows the electrical properties of the coil - I would imagine the acoustic output of the speaker probably has quite a few more resonances

Yes, but  generally the acoustical resonances will have some effect upon the impedance, as the voice coil impedance consists of both a static element and a back EMF from motion of the coil. Hence the acoustic impedance alters the back EMF to some degree by changing the motional loading of the coil and hence the measured |Z| and/or phase angle.

The attached plot shows the electrical impedance of a speaker with two mountings. Both demonstrate a strong resonance around 400 Hz, which is due, in my opinion, to acoustical resonance of the speaker and (in the second case) resonance of the combined speaker and mounting plate. The possibility of this being an LC resonance due to distributed capacitance can be discarded rather easily with a quick calculation of the required capacitance value.

One is the speaker frame resting on my measurement bench, and the second is the same speaker mounted in its normal mounting plate. The speaker is the internal device normally supplied with an Elecraft K2 ham radio transceiver and is attached to an L shaped metal plate that forms the top enclosure of the transceiver cabinet. The second measurement is the speaker mounted on the L panel but the panel is not installed in the transceiver (won't fit - my K2 has an optional 100 watt amplifier module where the speaker normally mounts) but is resting on the bench.

Even though the changes between the two mounting arrangements are not great, there's enough change in acoustic loading to be seen in the impedance measurement. I would expect a greater change in electrical impedance and resonant frequency if the speaker were mounted in the normal arrangement.

 

[krivx]

Interesting! This just shows the electrical properties of the coil - I would imagine the acoustic output of the speaker probably has quite a few more resonances

Yes, but  generally the acoustical resonances will have some effect upon the impedance, as the voice coil impedance consists of both a static element and a back EMF from motion of the coil. Hence the acoustic impedance alters the back EMF to some degree by changing the motional loading of the coil and hence the measured |Z| and/or phase angle.

This makes sense - I guess it depends on the measurement set up too (as usual). The back EMF will vary depending on what the test signal is and how the coil is driven. Do you get different impedance measurements with an earbud in your ear? Can you hear the test chirp signal?

[Monkeh]

They are being driven by what I am assuming is audiofoolery grade equipment.... it clearly states 600 ohm amp.

Creative? Audiophoolery? No. Just midrange hardware with the world's worst drivers, and absolutely no support. And a massive price tag for EAX! OMG EAX, SO REALISTIC, MUST HAVE EAX!