a couple of questions re: woofer roll of (upper ) and mc-2 info

[Nanook]

Looking for a technical discussion/calculation regarding natural woofer upper end roll off. What is the theory that explains this, what parameters effect this?

Also a blank call out to anyone with further information regarding the Musical Fidelity Mc-2s.

Thanks in advance.

[sreten]

Top end roll-off / response of bass/mid units is a very complex issue.

At mid frequencies the cone no longer moves as a piston.
You can get standing waves in the cone causing peaks.

Two main issues are :

1) The profile of the cone. A flat profile is axially very stiff and
goes high but has poor stiffness in the radial direction.
Curving the cone gives up some axial stiffness to increase
the radial stiffness.

2) Termination of the cone edge. When a wave hits the surround
the less energy that is reflected back into the cone the better.

Generally maximising bass performance causes problems in the
midrange, good midrange - usually because of the surround
causes some limitations in bass due to the damped surround.

For example generally foam surrounds are good for bass but not
so good for midrange, and PVC surrounds are good for midrange
but not so good for high excursion bass.

sreten.

[Tim Moorman]

Nanook,

Think of a woofer as nothing more than a motor governed by the usual electrical properties.

Many will argue that the mass of the cone (or lack of) is the primary determinant for extended woofer response. Not so. It simply sets the point at which inductance dominates the response.

Somewhere past the midband of a woofer's rising response, there is a point, Rmin, where the inductive reactance of the voice coil and the capacitive reactance of the mass are equal but opposite, cancelling each other out. Beyond this point, as frequency continues to increase, inductance dominates the circuit and forms the equivalent of a low pass filter, rolling off response.

Not much difference here from a coil's inductance being used in a crossover to roll off response.


Tim Moorman

[sreten]

Originally posted by Tim Moorman
Many will argue that the mass of the cone (or lack of) is the primary determinant for extended woofer response. Not so. It simply sets the point at which inductance dominates the response.
Tim Moorman

Hmmm...... this may be true of the lumped mass of the cone.

But simply put at higher frequencies when the effective mass
of the voice coil = the effective mass of the cone (cone effective
mass decreases with frequency above the point motion is no
longer pistonic ) this sets a fixed limit to the extension of the driver.

sreten.

[hunter audio]

1.

other than the cone profile and edge limitations by way of suspension part linearity mentioned above
(now a days reflected waves can be done away with easily )
the following will be of interest

2. voice coil iductance is one of the main limitation factor

higher the inductance more bass-ish and for a woofer which will have an appreciable x- max the vc will get larger and so will the inductance

3. larger inductance gives rise to - vc inductance non linearity which directly relates to the distortion factor of the drive unit

distortions are most audible in the mid band threrefore heavy inductors are normally used to suppress mid bands to woofers

there exists ways to have a large vc x- max without increasing the vc iductance dramatically _ to produce good wide band woofers

4. from personal experience

the material of vc wire , al is lighter than cu and the best choice for wide band

in pro audio al to cu gives a spl increase but does not have much effect on the cut off

where as i have seen woofer s with al vc s are better wide band woofers

5.
of course the rise time , moving mass , transience _ are other factors in this woofer versus a wide band woofer dibate

6.

cone material is of a critical factor ,

the mechanical structure of the cone material (coupled to a high loss surround) is critical

when cones can be produced to be absorbant that can absorb audible freqencies easily starting from around 100 hz to limit a wide band application of that woofer

suranjan

transducer design engineer

[sreten]

Originally posted by hunter audio
(now a days reflected waves can be done away with easily )

Its still a problem in rigid cones as far as I'm aware, that
is drivers using a very stiff cone material e.g. metal.

sreten.

[Svante]

Originally posted by Tim Moorman


Not much difference here from a coil's inductance being used in a crossover to roll off response.

Yes there is a big difference. Due to eddy currents, the voice coil inductance is lossy. The impedance is *not* w*L but rather w^n *L, where n typically is in the range 0.6..0.7.
This leads to a slope of a silly -4 dB/octave rather than the -6dB/octave that would be the case with the ideal inductor.
Some simulation software models this lossy voice coil inductance by a second L//R set of components, but this is only an approximation of the w^n behaviour.

Superimposed on the effect of the inductor is the cone break-up pattern, which in effect reduces the effective radiating area of the piston (cone), off-axis drop due to directivity, and also baffle step effects.
The baffle step is an *increase* towards higher frequencies, but in the transition region it can to some extent be compensated by the voice coil inductance yielding a more or less flat response and then when the baffle step reaches its plateau of +6dB, the level starts to drop.

So, in order to reach a smooth (electric-) filter-free roll-off, one has to take voice coil inductance (including lossiness), cone break-ups and the baffle step into consideration.

[Tim Moorman]

Svante,

The poster inquired about roll off on the high end of a woofer's bandwidth.

Inductance, lossy or not, is the dominant influence in this region, and essentially filters, and limits, the high end response.

Baffle step and cone break up, although valid concerns, are really quite different matters, governed by different forces.

Tim Moorman

[Tim Moorman]

Suranjan,

Agree. Had the opportunity to witness a Klippel test and print out on a low cost woofer I was modeling, and had used in a bass horn.

Aside from non-linearities in the suspension, inductance was the primary source of distortion.

Also, beyond about 70% xmax, many of the poorly made, long throw woofers available for sub/bass use exhibit double digit 2nd harmonic distortion.

The use of copper shorting rings seems to be the most effective way to keep inductance in check.

For a woofer to be extended on both ends of the spectrum, remain linear and articulate, and not cost a fortune, is the dream of all speaker builders. Designers too, eh?

Tim Moorman

[Svante]

Originally posted by Tim Moorman
Svante,

The poster inquired about roll off on the high end of a woofer's bandwidth.

Inductance, lossy or not, is the dominant influence in this region, and essentially filters, and limits, the high end response.

Sure, vc inductance is important. I just said it does not behave like normal inductors. This you can see in most impedance charts, like the one below. The curve is from Peerless web page and the driver is 830656. Overlaid on the web page is a semi-transparent simulation window, containing impedance (green dashed) response with (blue) and without (thick black) edge diffraction (0.2x0.6m baffle).
This particular impedance curve fits with a inductance n value of about 0.68, as in the model. There is a slope in the response that starts already at ~300 Hz, and at 2kHz the response starts to deviate from the model. In this case neither the model or the measurement has any baffle step (measurement is done in the wall of an anechoic chamber). Modelling baffle step/edge diffraction (blue curve) tilts the frequencies below 1000Hz, counteracting the effect of the voice coil inductance, and also affects the smoothness of the driver rolloff towards higher frequencies. Clearly directivity also plays a role, as seen in the red/green/thin black curves, as well as other mechanisms not modelled here (like cone breakups) in the region above 3kHz.

Baffle step and cone break up, although valid concerns, are really quite different matters, governed by different forces.

Tim Moorman [/B]

They are different, of course, but still important for the rolloff.