forward diaphragm displacement--have you heard of this?

[Bill F.]

Friends, Romans, countrymen,

I'm trying to track down information about a phenomenon I have heard about but never confirmed or understood. Can you help?

If, for example, you apply a low tone to a bass driver in a reflex or sealed enclosure that causes cone excursion of decent amplitude, I have heard that there is a certain amount of forward displacement of the cone; that is, the excursion midpoint moves forward of the resting point.

Confirmations? Explanations? Bewildered silence?

Gratzi!

Bill

 

[mrfeedback]

Assymetry ?

Hello Bill, yes I have seen systems where the bass cone displaced forward during strong bass passages.
I have never determined the cause but I have wondered about speaker port flow assymetry, or amplifier DC servo or something being tricked up.

Eric.

 

[Bill F.]

Thanks for the feedback, mrfeeback!

That's what I'm looking for--alternate theories. You see, I have a theory of my own that goes something like this:

The forward displacement is caused by the single-ended nature of air itself. At low frequency, a diaphragm is basically controlled by the air volume enclosed in the cabinet. When it compresses the volume on the rearward stroke, counterforce rises with excursion exponentially toward infinity. But as the diaphragm rarifies the enclosed volume on the outward stroke, the rate of return force increase falls off with excursion as internal pressure tends toward zero.

You could picture the cone as being held between two different springs; a stiffer one behind and a more compliant one in front. In a dynamic situation, I think this would cause the excursion midpoint to move forward as amplitude increases.

Does that make sense? I've never read a definitive explanation of this dynamic, which seems like it could really be an important one, affecting linearity and distortion. So I'm still fishing.

Bill

 

[kelticwizard]

I have seen the phenomenon you mention quite often. It comes from the asymmetrical magnetic system many speakers still have, and it is easily correctable-on the manufacturing line, not by the user-either one of two ways.

D.B. Keele calls the phenomenon "suck-in" or "suck-out" and says that the displacement can be either forward or back, in a vented or closed box.

I have only noticed inward displacement, and only in a reflex box. I'm not saying it does not occur in the other instances, only that this is how I have observed it.

The way the manufacturer can correct this is to:

A) Make the pole piece of such size so that there is just as much magnetic flux in front of the voice coil as behind it;

B) Put in a Faraday ring-otherwise called a "shorting ring".

Both methods are effective. The Faraday ring has the added effect of lowering the speaker's inductance.

In the "suck-in" I mentioned, as the frequency approaches Fb, the speaker quite visibly goes to the end of it's travel, and only moves from it's back end forward and back. Half the wave form becomes clipped off!

Incredibly, it doesn't actually sound that bad. When you clip off half a pure sine wave, you get high second-order harmonics, which are called "overtones". It actually comes out sounding pleasant. But needless to say, pleasant distortion or not what we are going for when we build an enclosure.

The fact that the result, while actually a rather drastic phenomenon, really doesn't sound that bad might well be the reason all but a few manufacturers kind of ignored the phenomenon until recently.

European makes dealt with the issue about 12 years or so, usually the more expensive makes. Peerless, with it's 260 model, was the first popular priced woofer that had a Faraday ring available in a speaker that was well under $100, to my knowledge. This had a lot to do with my becoming a Peerless fan.

Now, a great many woofer manufacturers are paying attention. When you see "Faraday ring", "shorting ring", or "symmetric magnetic drive" and talk of reducing second order harmonics, you know the manufacturer is taking care of this.

 

[AndrewJ]

The answer is simple and well understood. There is no need to invent new theories!!
The phenonemon is caused by non-linearity of the Bl value at high excursion that causes a d.c. force component from frequencies just above resonance to displace the coil towards the weaker flux region, which is usually outward. It will happen even if the driver is not in a box!
This is explained by Barlow (50th AES Convention collected papers) and by numerous papers since.
Good woofer design can largely control this effect. If you have seen this happen, the woofer is not so well designed or it is being driven to very large amplitude.

 

[Bill F.]

Enlightening!

Wiz, it's amazing to me that a cone could behave like this. From whence comes the force that overcomes the restoring force of the Vb and the spider?

Andrew, can you elaborate on the "d.c. force component," if that's the cause, or point me to (free) reading that goes more in-depth?

If it happens near resonance, is it a matter of signal losing it's grip on the VC as it leaves the gap, thereby allowing the moving mass to go balistic and create phase chaos upon reentry?

One more question: are you saying my idea about the Vb restoring force nonlinearities is incorrect, or simply that it isn't the major contributing factor in this phenomenon?

Bill

 

[kelticwizard]

I don't think the phenomenon is much due to imbalances from the compression of air in the cabinet. However, many posters have said that in a sealed box, there are non-linearities due to the compression of air in the box on the inward stroke. But I think that's a different story.


This issue was discussed before, so I am giving you the link. There is a link there to Lambda-makers of Stryke-which gives more info on this.
http://www.diyaudio.com/forums/showthread.php?threadid=1381&highlight=Faraday+ring

Oops! The Lambda link in that discussion is no longer valid. And it was so informative, too. Oh, well.

Don't feel bad if you tried to search for this previous thread and could not find it. I tried to search for it too but could not come up with it.

The only way I could find it was to write in "Faraday ring" on the diy.audio search. But of course, if you know about Faraday rings, you already know how to solve the problem, so the person who is new to the "suck-in" phenomenon is pretty well stuck.

As for your being amazed that a woofer can act this way-join the club. When I took my little home-made tone generator and garage sale test amp and tested my little bass reflex speaker for the first time, I could not believe what I was seeing. The phenomenon is very visible with a test generator, by the way. I was absolutely astonished.

I think this tends to happen a little more often with bass reflex models. I sometimes wonder if it is not the reason that sealed fans continue to look down on bass reflexes as being not high fidelity-because this "suck-in" situation is so widespread.

By their nature, the bass reflex model will always have a little more distortion than sealed. But I wonder if the sealed aficionados do not hear so many ported speakers with the "suck-in" phenomenon and figure that is the way that ALL bass reflexes sound. It really has only been in the last 10 years or less that speaker manufacturers have paid attention to "suck-in" in anything less than the top price ranges.

 

[SteveG]

By their nature, the bass reflex model will always have a little more distortion than sealed.

What? That's not the what I've heard! A well-designed bass reflex should have lower harmonic distortion than a sealed box, because of the damping of cone movement at the resonant frequency- more excursion = more distortion.
As far as transient response goes, well, the sealed and TL win hands down.
As far as the suck-in/out issue- I have seen it too, but never knew what caused it, and yes- it was on a ported box. Is this phenomenon only seen in bass drivers, or does it occur in midrange drivers too?
Steve

 

[JohnG]

I too have see this when testing a woofer at fairly high excursions. A asymmetrical magnetic field could easily cause this. You can think of it as being a poor rectifier, which will generate a DC force by rectifying part of the AC music signal.

 

[kelticwizard]

Steve:

First. let me just say that I am a proponent of the bass reflex system. In most bass applications, I think it is the way to go.

You are correct in what you say. At the tuning frequency, the average bass reflex will require only about one fourth the cone movement of the closed box producing the same SPL. If you work it out mathematically, this amounts to little if you cut the bass off above 100 Hz or so. But in a system where the bass speaker will be playing up through the midrange, the midrange in the sealed system will becme muddied by the extra cone excursion the sealed system requires. It is called Doppler distortion.

I will give you the recently late, (and VERY lamented) Paul Klipsch's formula for calculating Doppler distortion if you want. It is from a couple of articles he wrote for the Jounal Of the Audio Engineering Society in the early seventies. You might consider getting those articles off the Jounal website though, because Paul Klipsch, in plain language, cuts through A LOT of what good sound reproduction is all about. Five bucks each for 2 articles. Worth it.

As for the transient response. Okay, we agree. We both know that the bass reflex has a degree of "overshoot" that the sealed does not have. How great this overshoot is depends on the tyupe of bass reflex it is-the smaller the box volume compared to Vas, the less it will be.

This "overshoot" can be shown either by displaying the waveform on a graph OR on a spectrum analyzer as separate frequencies. Just as a square wave can be shown as a series of pure tones on a spectrum analyzer, so too can just about any type of distortion, of which "overshoot" is one, be shown the same way.

For any given ratio of box volume to Vas, (Vas/Vb), the closed box will have less "overshoot", therefore less distoriton. There are some bass reflexes whose Vb is so small compared to Vas that they have less "overshoot" than some sealed systems whose box volume is much greater, comparatively, to Vas. But for equal ratios of Vas/Vb, the sealed wins out every time.

Which is why you won't often see a bass reflex midrange even though such a driver would deliver less distortion in it's higher ranges.

In other words, essentially you and I are in agreement. I was only talking about one certain kind of distortion.

 

[Bill F.]

Destorshin

Of course in the final analysis, if it adds or subtracts--yea even a jot or a tittle--from the signal waveform, it's distortion! That's why I think it's funny to differentiate between distortion and transient response, except for academic purposes.

Getting back to the suckout thing, it seems to me that ported alignments might be more susceptable to them because a DC force, even if only temporary, will unload through the port. A sealed alignment would preserve the Vb air spring in this situation.

Bill

 

[SteveG]

Of course in the final analysis, if it adds or subtracts--yea even a jot or a tittle--from the signal waveform, it's distortion! That's why I think it's funny to differentiate between distortion and transient response, except for academic purposes.

yes distortion in any form is unwanted. That's a given. But, I think that phase distortion is one thing that is very often overlooked, and we will be hearing more about it in the future (a similar discussion came up when talking about SACD Vs. PCM 16/44.1.-see the mp3 sounds better thread) Transient response (or the lack thereof) in a speaker system is a form of phase distortion, correct? In this case, it isn't adding or subtracting- it's moving the signal to a different time. I can clearly hear it in my MTM's, which are tuned a little lower than they should be to get more extension.
It is amazing the effect that the port can have on driver excursion- if you set your signal generator to 38.2hz on my system, and it looks as if the woofers are standing completely still when seen from the side! I still prefer the "punch" that I used to get with an old set of MTM closed boxes with 10" woofers. Of course the midrange was lousy.
Doppler effect is an interesting thing. I have been considering this quite a bit in my next design. I have concluded that I want to negate it by using multiple drivers in a multi-way triamped system. While two ways are easy, I think 3 ways may have them beat for the just the effects that you mentioned where the woofer modulates the midrange. I would like to try a TL enclosure to see if they're all that everyone says they can be.
Bill F.
You nailed down what I was thinking- couldn't figure out how to describe it. Since the box dosen't load a woofer below Fb, the recification effect could produce a very low frequency signal that would not be damped by a bass reflex as it would be in a closed box design... interesting.
Kelticwizard,
Yes, I would like to see the formula, if at all possible.
Thanks.
Steve

 

[Bill F.]

I think that phase distortion is one thing that is very often overlooked . . . . In this case, it isn't adding or subtracting- it's moving the signal to a different time.

Amen. And the thing that makes it so egregious is that a the recipe for a waveform is equal parts pressure and time--stirred, not shaken. Change the timing of certain bands, and they will superimpose where they shouldn't--you've wacked the waveform, as the ear perceives it.

(Obviously, I'm a stickler for phase punctuality )

Bill