Speaker Sensitivity (1 Viewer)

[Dustin B]

I've read about and since regurgitated how a speakers sensivity relates to input power and spl. However, I started wondering today how meaningful a speakers sensitivity rating is. Is it an average across the entire frequency range of the speaker? Or does the sensivity of the speaker not change with varying freqeuncy?

Ie will a speaker's sensitivity be lower with bass frequencies than with higher frequencies or the same for both?

 

[Duncan Barth]

I'm new at this whole speaker building gig, but here's my understanding of it:

A speakers sensitivity will generally vary across its frequency range (electrostatic panels and planer ribbons seem to very a lot less) ... sometimes by quite a bit. The one number you get from the manufacturer is more or less an average.

A good example of this is to go to parts express and look up some of the drivers they sell. They usually have a pdf spec sheet that goes with them, with graphs showing how the parameters vary.

 

[Patrick Sun]

A speaker's rated sensitivity is more of a measure of how much SPL it will produce in its "normal" operating frequency range, but you have to take into account the overall frequency response of the total speaker as well. If the speaker is well designed, its frequency response will be +/- 3dB from 40Hz-20KHz (with the bottom octave filled in with a sub or side-firing woofers).

The normal overall sensitivity of the speaker will most likely depend on the midwoofer drivers because they tend to be less sensitive (consume more power), with the tweeters being shelved down with a bit of padding to equalize the sensitivities between the 2 drivers to get that +/- "flat" frequency response. So it's critical to find a well-behaved midwoofer that's reasonably flat in its operating range so you don't have to worry about the varying sensitivity of the overall speaker.

This is why some designs feature 2 midwoofers, in an effort to produce a speaker with a higher sensitivity rating because the extra midwoofer will theoretically add 3dB of SPL if wired in parallel with the other midwoofer, but at the expense of requiring an amp that can drive that larger load with a smaller impedance value (due to the parallel configuration of the midwoofers). But it's trick to get the integration of the 2 midwoofers with the tweeter, and get the polar response of the off-axis SPL output to sound good.

 

[Ron Shaw]

2 drivers in parallel results in an increase of 6db. Voltage in a parallel circuit remains constant ( assuming both drivers are the same, i.e. same impedance), current requirements will double. Basically you now have a voltage increase of 2. Using ohms law you can find that when you increase amplitude by 2 (double the voltage), your power increases by 6db. An increase of 1.414 (voltage) results in a 3db increase.

 

[ThomasW]

Dustin

Mfgrs usually measure the speaker at a specific frequency. And usually the frequency is chosen because that's where the speaker will have the highest output. As a result efficiency ratings are frequently not representitive of the output of the speaker over the full bandwidth of the driver.

 

[Ron Shaw]

All the driver manufacturers Im familiar with use an average value across the recommended range for the driver. If you find a manufacturer the uses a single frequency for their specifications, I would pass, and use a more ethical one.

 

[Patrick Sun]

Ack, you get 3dB acoustical AND 3dB electrical only if the amp can supply the necessary current given the lowered impedance. That's what I forgot. Thanks for the correction. But in everyday situation, plus baffle step, you can expect roughly 3dB of SPL output by parallelling the midwoofers.