Limiting Factors for high-frequency extension

[Nappylady]

It seems like there are really only two limiting factors on high-frequency extension of a driver: the inductance in the motor structure, which raises impedance with frequency, and cone breakup, which causes irregularities in frequency response and dispersion above the breakup frequency.

1. Looking at the impedance curve of a driver, I can see that a driver will pull less current for higher frequencies because its impedance at those frequencies is higher. Can I simply increase the voltage output on my amplifier for those frequencies to compensate for the rise in impedance? (Providing that I'm not driving the amp to clip, of course!)

2. I see lots of things done to handle breakup modes, like using stiffer cone materials (metals like Aluminum or Magnesium) or driving more of the cone with a larger voice coil. I have an idea about using multiple motors & voice coils to drive a flat diaphragm from the edges, and I wonder if this could be effective, and brings me to question #3:

3. Mms is *not* a factor in HF extension, right? My intuition tells me that it could be, but the more I think about it, the more I think that it will only make a difference to sensitivity, but not to frequency response. Is this right?

[holger honda]

maybe one needs to add that the size,weight and material of the cone will also limit the upper freqs.
Im working alot with compression drivers and even a 2 inch diaphragm compared to a 1 inch will have a harder time in the upper freqs.And thats just a little flat thing of super rigid titanium....I can just imagine trying a 6.5 inch...sluggish.

[soongsc]

Originally posted by Nappylady
It seems like there are really only two limiting factors on high-frequency extension of a driver: the inductance in the motor structure, which raises impedance with frequency, and cone breakup, which causes irregularities in frequency response and dispersion above the breakup frequency.

1. Looking at the impedance curve of a driver, I can see that a driver will pull less current for higher frequencies because its impedance at those frequencies is higher. Can I simply increase the voltage output on my amplifier for those frequencies to compensate for the rise in impedance? (Providing that I'm not driving the amp to clip, of course!)

2. I see lots of things done to handle breakup modes, like using stiffer cone materials (metals like Aluminum or Magnesium) or driving more of the cone with a larger voice coil. I have an idea about using multiple motors & voice coils to drive a flat diaphragm from the edges, and I wonder if this could be effective, and brings me to question #3:

3. Mms is *not* a factor in HF extension, right? My intuition tells me that it could be, but the more I think about it, the more I think that it will only make a difference to sensitivity, but not to frequency response. Is this right?

Just turning the amp up does not compensate for impedance gain. You really need to have gain that varies with frequency to compensate for the impedance. You can do this by active compensation or passive compensation. Passive compensation sacrifices efficiency.

[Nappylady]

soongsc, yes that is what I meant :-) gain that increases with frequency to compensate for the rise in impedance.

(my thinking is that) because the same amount of current is being drawn, the extra gain isn't going to blow out voice coils faster than playing low frequencies; so as long as I keep the amp lower than clipping for those frequencies, I should be just fine.

honda, I'm not sure that's right. The right side of my brain says that mass will have an effect on HF extension, but my left says no, that doesn't make any sense. An increase in mass should make the driver less efficient across all frequencies, not just more "sluggish".

Moving from a 1" to a 2" diaphragm introduces a large increase in inductance and breakup modes, I'm sure. Do you think maybe it's those two factors you're seeing?

[soongsc]

Sounds like you are going to drive certain drivers beyond their original intent?

Greater the impedance change, the lower efficiency you will end up with after compensation if you do it passively. It would be neat if you can do something software/firmware programmable.

[Nappylady]

Passive? No way.

I'm actually thinking of a simple filter using an op-amp with a pole and two zeroes (or a zero and two poles) in the feedback function... Don't ask me; I don't know calculus!

[hermanv]

Just mussings not intended as real knowledge.

Don't most drivers fall off at 12dB/octave so 4 times the power for another octave?

Also I believe that it's not the coil that stops moving due to inductance ('course inductance does reduces movement) but the cone no longer follows, even the metal dome tweeters flex.

There are several super tweeters on the market and the diamond tweeters reach 100KHz (not cheap of course).

This is interesting, my wife can hear those ultrasonic burglar alarms (33+KHz?) and easily hears bats ("What's that chirping noise?")

Where do you get source material?

[BassAwdyO]

I wonder what kind of transducers are used for ultrasound scans. I bet they can go pretty high

[DcibeL]

Originally posted by BassAwdyO
I wonder what kind of transducers are used for ultrasound scans. I bet they can go pretty high

Something perhaps made by this company? I would assume it wouldn't be such a hard task to design a transducer for ultrasound application, since they only need to transmit a single frequency, and therefore can have a very short bandwidth of frequencies they can produce.

[simon5]

It's a good idea NappyLady, but you'll need to consider off-axis response or your speaker will sound nice only in a small sweet spot...