Voice Coil Inductance (Le) and Transient Responce

[amo]

Hi-

An audio engineer that works for a large loudspeaker importer here in the US has recently explained to me that if I want to know the transient responce of the driver it self (without reguard to enclosure design that is), I should look at the Le or Inductance of the voice coil. My question is how much difference does this exactly make? For example, one tweeter I am looking at has Le of .01 and another of .07. Does this play a large role, especially since an enclosure for the tweeter is vertually non-existant (or has no effect at least in my case). When we speak about sub woofers, the difference in Le is quite large. For example, many are areound 2.35. However, there are some that are below .5! So will the 2.35 "swallow up" sonic details (assuming closed box design)?

Thank you for your insight!

[Nappylady]

A coil is an inductor; that is to say, that as frequency rises, impedance increases.

Yes, then, higher inductance means that higher-frequency harmonics will get cut off.

As always, we go to the square wave as our first example--woofers restrict the upper frequencies, rounding off the edges and creating troughs in the tops and bottoms of the square wave.

When fed a sine wave, the woofer will move slowly and not adjust quickly--like a sumo wrestler trying to do ballet--consequently, there's phasing problem (group delay, except now in context of a driver instead of a crossover) as the woofer can't keep up with the sine wave; in an extreme case, it would actually look like the integral of the sine wave, because it is shifted so much. If the sine wave ends abruptly at 0, the woofer continues moving for a short time afterwards while the suspension pulls the poor thing back into normal position.

Motion feedback with accelerometers helps with this immensely, but the bottom line is, you would need a driver with infinite power handling, an amp with infinite power, and a perfect, delay-free accelerometer setup to get a single driver to have perfect transient response. At that point, you'd probably end up with perfect frequency response as well, and of course, perfect phasing.

I suspect that, with some calculus, a computer could measure, in excruciating detail, the characteristics of a speaker setup, then perform a transformation, or set of transformations, in order to acheive as close to perfect transient response, phase response, and frequency response--perhaps 0.00001% THD, flat from 20-20khz +/- .01 dB, and phase correct +/- 1 degree--and with todays computers, it could be done in (almost) realtime. (I think you'd have to delay things by about 50 ms, but unless you're beatmatching, you won't even notice. 50 ms when you pause the CD player is NOTHING. )

Of course, I'm not the one to write the software... but wouldn't that be cool?

[Nappylady]

Um... that was a very meandering reply because it's late and I'm tired and I'm sorry about the hijack.

In short: it's another compromise. If you go with lower inductance woofers, you'll probably sacrifice low-end response; but with higher-inductance woofers, you lose higher-end response because of the high inductance. Splitting the system is, of course, the only way to get things right.

[phase_accurate]

Le is in fact one of the factors determining transient behaviour but by far not the only one.
The negative effect of Le can be taken care of by using current feed instead of voltage feed when using active designs.

Regards

Charles

[kelticwizard]

Originally posted by Nappylady
In short: it's another compromise. If you go with lower inductance woofers, you'll probably sacrifice low-end response; but with higher-inductance woofers, you lose higher-end response because of the high inductance. Splitting the system is, of course, the only way to get things right.

How does an aluminum "short circuiting" ring in the magnetic system fit into this? It is also called a Faraday ring.

I know from experience that the ring prevents "suck-in" in a reflex box near the box tuning frequency. This is a process where the voice coil moves to the extreme end of it's travel and sits there, essentially shearing off half the wave form. The "short circuiting" ring prevents this. Another way to prevent this "suck-in" is to extend the pole piece to provide balance in the magnetic circuit.

Have heard some talk that this "short circuiting" ring also cuts down on inductance-but I have not heard anything about it limiting low-end performance. The extended pole piece method does not limit inductance, from what I have read.

I am not an engineer. I just witnessed the "suck-in" phenomenon when either of these two methods are not employed. I also note that good quality woofers increasingly employ one of these two techniques.

Is this "short circuiting" ring what you mean by "splitting the system"?

[capslock]

The ring is roughly 20x more conductive than the iron alloy used for the pole piece. When a current flows in the VC, it will induced a mirror current in the ring which is shorted. This effectively keeps the AC magnetic field of the VC from entering the iron, so that it cannot act as an inductor core. Remember, an iron core "amplifies" the field of the coil, thus increasing the inductance.

[kelticwizard]

Thanks, Capslock. So can it then be said that the "short circuiting" ring cuts down on inductance while not necessarily limiting low frequency response?

[capslock]

It isn't that a high inductance generates any extra LF response, rather it attenuates the highs, thereby producing a RELATIVE bass boost.

Cheers,

Eric

[Circlotron]

I bet if the signal used to measure the coil was of several volts magnitude and say 30-40 Hz, the inductance and reactance of the coil would measure way differently it were moving or held still.

[amo]

Thank you for the insight! It has helped me with driver selection for this particular project, and it has been very educational as well. However, I am still at a loss on tweeters. Keeping in mind the above information, please consider the following two HF drivers. It must be said that the frequency responce curves for both drivers are very similar.

Tweeter A

Nom. power handling 15 W
Max. power handling 150 W
Sensitivity 94 dB (2.8V/1m)
Cone Titanium
Surround Foam
Nom. impedance 4 §Ù
DC resistance 4.5 §Ù
VC diameter 20 mm
VC height 2 mm
Former Aluminum
Layers 2
Wire Copper
Inductance 0.32 mH
Gap height 2 mm
Net weight 0.14 kg
Fs 1379 Hz
Qts 0.8
Qes 1.56
Qms 1.64
Re 5.84 §Ù


Tweeter B

Nom. power handling 15 W
Max. power handling 150 W
Sensitivity 95 dB (2.8V/1m)
Cone Titanium
Surround Coated Foam
Nom. impedance 8 §Ù
DC resistance 6 §Ù
VC diameter 20 mm
VC height 2.2 mm
Former Aluminum
Layers 2
Wire Copper
Inductance 0.08 mH
Magnet diameter x height 51 x 7 mm x mm
Magnet weight 69 g
Flux density 1.98 T
Gap height 2 mm
Net weight 0.3 kg
Fs 1058.78 Hz
Qts 0.51
Qes .77
Qms 1.51
Re 6 §Ù
Res 11.7 §Ù

Now I should say that driver B costs about 5 times the price of driver A!!!. Driver A is also prefered due to its smaller footprint (at least for this project). The with all these facts, will driver B really produce much more detail? Max SPL is not an issue here, only detail, transient responce, and lack of distortion are. Both drivers will be crossed over at about 2.5 Khz, and will get 100 W of power.

Thank you for your help!!!!