Q: Why is high Le bad for hi-fi subwoofer?

[maximiliano]

I heard that high VC inductance is not good to have for a subwoofer for music application (maybe also for HT). Whis is that so?

[Pete Schumacher ®]

> I heard that high VC inductance is not good
> to have for a subwoofer for music
> application (maybe also for HT). Whis is
> that so?

High inductance can be an issue when it comes to bandwidth and some types of distortion. But with the limited bandwidth of a sub, it shouldn't be an issue all by itself, especially if you're limiting it to the first couple octaves.

[maximiliano]

> High inductance can be an issue when it
> comes to bandwidth and some types of
> distortion. But with the limited bandwidth
> of a sub, it shouldn't be an issue all by
> itself, especially if you're limiting it to
> the first couple octaves.

Thanks for the response, Pete. If a sub is crossed at 80 or 100 Hz, signal up to 200 Hz may be well audible from the sub. Won't high Le affect performance in this case?

BTW, why do some people say it's bad to have high Le in a subwoofer driver? Usually a subwoofer is used without any passive XO. I don't understand why high Le is bad.

[daryl]

Provided Link: Daryl's Speaker Stuff. (http://www.villagephotos.com/pubgallery.asp?id_=144711)


Le causes trouble in several ways.

1. Flux modulation.

Le is not the cause but and indicator of this.

You have two magnetic fields in a driver, that of the magnet and that of the voice coil.

The magnetic field of the voice coil modulates the magnetic field (flux) of the magnet causing even order distortion.

You want the magnetic field of the voice coil to be small relative to that of the magnet so that it will have only a small effect.

Things which increase Le also increase flux modulation like using a small magnet and an eight layer voice coil instead of a large magnet and a two layer voice coil or making the voice coil longer to increase x-max.

2. Le isn't linear.

If the reactance of Le is signifigant it will cause distortion because of it's non-linearity.

The voice coil is an inductor with an iron core.

The iron is excellent at conducting the steady flux of the magnet but not a suitable material for an inductor core, even at subwoofer frequencies.

This is why Le changes with frequency and is lossy, it's the hysteresis of the iron in the motor.

3. Inductance modulation.

As the voice coil moves through it's excursion range Le often changes with position due to the coil moving off the end of the pole-piece.

Driver designs with hyperextended pole-pieces (not extended pole) minimize this effect.

4. Peaking and rolloff from Le.

Rolloff is relatively benign but Le also resonates with Mms and causes a peak just above Fs/Ftc that most speaker software doesn't even show you.

This causes the result to be more of a peak than the flat extended response you were shooting for.

There are a lot of factors but a good rule of thumb is Re/Le should be more than 2000 for subwoofers and 8000 for woofers.

That's not always though, if you have measurements that say otherwise then so be it.

> I heard that high VC inductance is not good
> to have for a subwoofer for music
> application (maybe also for HT). Whis is
> that so?

[billfitzmaurice]

> BTW, why do some people say it's bad to have
> high Le in a subwoofer driver? Usually a
> subwoofer is used without any passive XO. I
> don't understand why high Le is bad.

As Pete said, it isn't, within reason. High Le will cause a loss of response in the midrange for sure and possibly in the midbass, but since subs don't go that high it's a rather moot point.

[unknownuser]

Maybe it's the 50-60hz hump in response some have mentioned getting with a high Le driver.

> BTW, why do some people say it's bad to have
> high Le in a subwoofer driver? Usually a
> subwoofer is used without any passive XO. I
> don't understand why high Le is bad.



(Originally posted by: Frank)

[billfitzmaurice]

> Maybe it's the 50-60hz hump in response some
> have mentioned getting with a high Le
> driver.

That's where 'within reason' comes into play. 2mH is reasonable, 10mH isn't.

[JRT]

> That's where 'within reason' comes into
> play. 2mH is reasonable, 10mH isn't.

To your point, the Peerless 830500 12" XLS is a pretty good subwoofer below ~125_Hz, and the datasheet lists its Le at 4.2_mH. Perhaps no too coincidentally, impedance begins rising above that range.

[jonmarsh]

> Le causes trouble in several ways.

> 1. Flux modulation.

> Le is not the cause but and indicator of
> this.

> You have two magnetic fields in a driver,
> that of the magnet and that of the voice
> coil.

> The magnetic field of the voice coil
> modulates the magnetic field (flux) of the
> magnet causing even order distortion.

> You want the magnetic field of the voice
> coil to be small relative to that of the
> magnet so that it will have only a small
> effect.

> Things which increase Le also increase flux
> modulation like using a small magnet and an
> eight layer voice coil instead of a large
> magnet and a two layer voice coil or making
> the voice coil longer to increase x-max.

> 2. Le isn't linear.

> If the reactance of Le is signifigant it
> will cause distortion because of it's
> non-linearity.

> The voice coil is an inductor with an iron
> core.

> The iron is excellent at conducting the
> steady flux of the magnet but not a suitable
> material for an inductor core, even at
> subwoofer frequencies.

> This is why Le changes with frequency and is
> lossy, it's the hysteresis of the iron in
> the motor.

> 3. Inductance modulation.

> As the voice coil moves through it's
> excursion range Le often changes with
> position due to the coil moving off the end
> of the pole-piece.

> Driver designs with hyperextended
> pole-pieces (not extended pole) minimize
> this effect.

> 4. Peaking and rolloff from Le.

> Rolloff is relatively benign but Le also
> resonates with Mms and causes a peak just
> above Fs/Ftc that most speaker software
> doesn't even show you.

> This causes the result to be more of a peak
> than the flat extended response you were
> shooting for.

> There are a lot of factors but a good rule
> of thumb is Re/Le should be more than 2000
> for subwoofers and 8000 for woofers.

> That's not always though, if you have
> measurements that say otherwise then so be
> it.

Good post, Daryl. High Le drivers is a major reason some folks don't like long throw woofers/subwoofers, because, quite frankly, they don't sound as good. They may sound "good enough", especially if rumbling LF is your main concern. But inductivity modulation raises distortion when it SHOULD be falling due to smaller cone excursion with increasing frequency. The inductance causes roll off in the response, and makes integration with crossover points like 75 Hz or 100 Hz more challenging- I can get a high Le driver crossover to integrate OK at 80 or 100Hz, but it usually requires EQ separate on the driver to correct the response in the 100 Hz to 150Hz area so the actual response follows the acoustic target.

Last, since this is a PE supported forum, go look at those RS sub drivers- all have quite low Le, and the overall performance in a variety of applications is excellent because of that.

BTW, I've worked with and own a number of high Le drivers (the original Stryke HE-15, the Blueprint series (12's and 15's), as well low Le drivers like the RS subs and Aurasound woofers and sub drivers, and Ciare (Italian) pro drivers, so I'm relating this from personal experience.

[Jethro]

> Good post, Daryl. High Le drivers is a major
> reason some folks don't like long throw
> woofers/subwoofers, because, quite frankly,
> they don't sound as good. They may sound
> "good enough", especially if
> rumbling LF is your main concern. But
> inductivity modulation raises distortion
> when it SHOULD be falling due to smaller
> cone excursion with increasing frequency.
> The inductance causes roll off in the
> response, and makes integration with
> crossover points like 75 Hz or 100 Hz more
> challenging- I can get a high Le driver
> crossover to integrate OK at 80 or 100Hz,
> but it usually requires EQ separate on the
> driver to correct the response in the 100 Hz
> to 150Hz area so the actual response follows
> the acoustic target.

> Last, since this is a PE supported forum, go
> look at those RS sub drivers- all have quite
> low Le, and the overall performance in a
> variety of applications is excellent because
> of that.

> BTW, I've worked with and own a number of
> high Le drivers (the original Stryke HE-15,
> the Blueprint series (12's and 15's), as
> well low Le drivers like the RS subs and
> Aurasound woofers and sub drivers, and Ciare
> (Italian) pro drivers, so I'm relating this
> from personal experience.

I agree. I went from a TC built sub (Elemental Designs 12O.14) with a measured Le of 5.5 mH, to an RS 12HF (measured Le of .95 mH) -- and there was a night and day difference in the ability to integrate smoothly with the rest of my system. I also noticed the response peak right above Fc with the TC sub -- which is obviously not present with the HF sub, despite being in the same cabinet, using the same amp, and having an almost identical sealed alignment (just under .707 Qtc, Fc just under 40 Hz).