speaker impedance at very high frequencies

[kwantam]

Hello all,

I'm looking for general information on speaker/crossover network input impedance at high frequencies, around 1MHz.

I'm aware that different crossover topologies will behave differently, and perhaps this question is just too wide open to have any sensible answer at all. Nevertheless, I'm interested in whatever comments you can give on the matter.

So: at frequencies in the MHz neighborhood, do speakers generally still look inductive? Do certain classes of speakers look more so than others?

I'm assuming that electrostatic speakers look capacitive, since they're, um, capacitors. So let's neglect those and concentrate on dynamic speakers.

Any comments would be appreciated.

[AllenB]

You may need to consider the effect of interwinding capacitance between the inductor turns as well as voice coil turns. I'd be measuring if possible.

[jcx]

inductors, including voice coils will have "Self Resonant Frequencies" where the parasitic C resonates with the coil L

for iron core inductors the 1st SRF can be low 100s kHz, probably many MHz for single layer voice coils

for most systems Cable C in particular will be more important to a amp, longer cables may begin to show transmission line 1/4 wave resonance over a few MHz

the only time driver/XO MHz Z si seen directly is if the amps is mounted on the back of the speaker

ESL intended for use with typical audio power amps have step up transformers, leakage L resonance with the panel C gives a Z peak a few octaves above audio – beyond that the amp mostly sees the xfmr leakage L, any SRF peaks

[kwantam]

AllenB, the inter-winding cap is what I was thinking about as well. As you say, measurement is a sure way to an answer.

jcx, yes, we're talking about the same thing: if we model it as a simple inductor with parasitic shunt cap, above the resonant frequency the impedance looks capacitive. Even with a more realistic speaker model, a shunt cap at the input will eventually dominate the impedance. You make a good point about the cable capacitance; that'll surely be a factor.

Interesting info on electrostatics; thanks very much.

[woody]

And of cource a peizo will just look like what it is a capacitive load.

[whgeiger]

Originally Posted by kwantam

Hello all,

I'm looking for general information on speaker/crossover network input impedance at high frequencies, around 1MHz.

I'm aware that different crossover topologies will behave differently, and perhaps this question is just too wide open to have any sensible answer at all. Nevertheless, I'm interested in whatever comments you can give on the matter.

So: at frequencies in the MHz neighborhood, do speakers generally still look inductive? Do certain classes of speakers look more so than others?

I'm assuming that electrostatic speakers look capacitive, since they're, um, capacitors. So let's neglect those and concentrate on dynamic speakers.

Any comments would be appreciated.

As the diaphragm will not move at these freqiencies, some of the energy will be radiated from the voice coil (like an atenna) and the remainder will be turned into heat.

Regards,

WHG