Damping factor and shorting rings

[peniku8]

I don‘t know how the shorting rings in a driver work mechanically, but as I was thinking about damping factor and that long cable runs can get problematic with low impedance systems, my head told me that the shorting rings also play a role here.

Do they have an influence on the damping of the system?

Is the impact big enough to neglect the amplifier-cable-system damping factor if your drivers have shorting rings?

[SME]

I will try to explain as best I can here.

Shorting rings don't act mechanically, at least not directly.  They magnetically interact with the voice coil to reduce and/or linearize inductance through the driver's stroke.

The voice coil is a natural inductor.  Inductors store and release energy via the magnetic field in their proximity.  For a straight wire with pure resistance, the current responds in perfect lock-step to changes in voltage.  If an inductor is subject to a sudden increase in voltage, however, some time and energy are required for the current to "spin up" the magnetic field.  At that point, if the voltage is suddenly cut to zero then the current flow continues for some time while the magnetic field "spins down" and releases stored energy.  Hence, rapid fluctuations in voltage tend to be smoothed out in current, e.g. high frequencies are reduced.  This is a major cause of high frequency loss and sometimes "humping" in a speaker driver's response.

At the same time, the inductance itself is likely to vary, not just with frequency but with changes in instantaneous current and/or driver stroke.  This is *non-linear* behavior, which causes distortion, including inter-modulation distortion, which may be particularly undesirable.  Though the linear aspects can also be degrading if not precisely corrected with EQ.  The response "humping" alters the spectral balance and likely imparts a non-neutral characteristic sound.

A shorting ring is made from a material that is both magnetically and electrically active.  The magnetic field generated by the current induces current flow within the ring.  The ring "shorts" this current to the rest of the driver assembly, allowing some of the energy in the magnetic field to be transferred to the shorting ring and dissipated as heat instead of stored.  If designed correctly, this effectively reduces the inductance of the coil), and depending on the position of the rings vs. the voice coil, may also keep inductance from fluctuating as much throughout the stroke.

The relationship between inductance and damping is via the electrical impedance.  Impedance is essentially a 2-dimensional quantity which can be described in terms of a pair of parameters: either *magnitude* and *phase angle* or *real* (resistive) and *imaginary* (reactive).  Inter-conversion is possible via basic trigonometry; see the "Complex plane".  Damping is a property of the resistive / dissipative (non-energy storing) component of impedance.  Pure inductance and capacitance both contribute only to the reactive (energy storing) component of impedance.  Speakers using a composite electrical circuit that has effective resistances, inductances, and capacitance contributed by several different factors including the mechanical and acoustic properties of the system.  So needless to say, inductance and "damping factor" both contribute to the system behavior in a way that's not simple to describe.

To answer your last question: No.  I mean, if the resistance of your speaker wire is high enough be a problem in the absence of shorting rings, then shorting rings probably won't fix that problem.

[peniku8]

Thanks for the great explanation! So in short: the shorting ring magnetically opposes the (self-)induced field of the voice coil to reduce apparent induction by basically converting that energy to heat?

If I understood that correctly, the shorting rings lower the voltage sensitivity as well?

And the IPAL system works on a similar basis, just on the electric side, basically eliminating anomalies produced by induced current?

[SME]

10 hours ago, peniku8 said:

Thanks for the great explanation! So in short: the shorting ring magnetically opposes the (self-)induced field of the voice coil to reduce apparent induction by basically converting that energy to heat?

I'm not 100% sure on whether or not the energy that would have been stored in the field is dissipated, but I believe it is.  I'm guessing it's typically a small amount, which may reduce the driver power efficiency slightly but I don't know.

10 hours ago, peniku8 said:

If I understood that correctly, the shorting rings lower the voltage sensitivity as well?

 

No.  Shorting rings actually *increase* voltage sensitivity by allowing higher current, specifically for high frequencies.  Higher coil inductance increases the lag time for current to respond to changes in voltage.  The higher the frequency, the faster voltage is oscillating and the less time there is for the current to get up to speed before the voltage reverses again.

12 hours ago, peniku8 said:

And the IPAL system works on a similar basis, just on the electric side, basically eliminating anomalies produced by induced current?

I'm not sure which aspect of the IPAL system you are referring to here.  For amps that can maintain consistent power output into very low driver impedances, voltage sensitivity (such as inductance causes for high frequencies) is less likely to be a limitation, but that doesn't correct all the problems that inductance causes.  EQ / signal shaping can be used to compensate for high sensitivity frequency droop, as is often done in passive crossovers, but the correction is likely imperfect and may allow some "sound signature" to persist.  Furthermore, inductance does not behave especially linearly with medium-to-high signal levels, and this can cause significant compression and distortion effects.  Well-designed shorting rings directly reduce and linearize inductance and so help with all of these problems.

[kipman725]

18Sound have some interesting drivers that have little/no inductive impedance rise using a technology they call AIC:

http://www.eighteensound.com/media/W1siZiIsIjIwMTgvMDkvMTAvMTFfNDZfMDNfNzU1X0FJQy5wZGYiXV0/AIC.pdf

http://www.eighteensound.com/en/products/lf-driver/10-0/8/10NMBA520

I think KV2 might be using their drivers.

[SME]

Yes.  Their solution is very interesting, but it's not implemented on many drivers.  I wonder why not?

Acoustic Elegance designs their woofers with a long copper sleeve in very close proximity to the voice coil which dramatically cuts inductance in their drivers.  AE also claims that their sleeves improve thermal performance, but I have not seen independent verification of this claim.  Some other sub manufacturers (Funk audio, maybe others?) also use full sleeves, albeit not installed as close to the voice coil as in AE's designs but close enough to yield impressive performance and linearity.

[kipman725]

Copper sleeves are addressed in the AIC paper as a superior version of the shorting ring as they allow a smaller gap size increasing gap flux (and presumably improving thermals as AE claim). 

The AIC might just be more expensive to manufacture and hard to market.

 

 

[peniku8]

I was comparing it to the linearization effects of the IPAL system. Afaik the system monitors driver specs to linearize thermal effects for example, and the shorting ring would just be the inductance equivalent then. I could be wrong and the system is less advanced than I think. 
Just that thermal effects only arise above a certain output volume and inductance is always present.

The real reason for the ipal amp seems to be the driver thou. In order to meet the design goals, the driver ended up with a super low impedance, which needed the custom amp and requires to build an active cabinet.

[SME]

Interesting.  I wonder how fine-grained the IPAL system is here?  Does it adjust EQ in real-time or just gain?  If it adjusts EQ, how tight is it?  Do the sound characteristics noticeably wander through a live performance?   It's not too hard to monitor DC resistance and adjust gain in real-time, but that may not be good enough.

I expect that thermal changes manifest over seconds to minutes or hours.   If we assume all of the thermal effect is to raise the coil DC resistance, then the nature of the non-linearity is such that we can understand things in terms of linear response that changes with time.  This is not possible for inductance because inductance can fluctuate much more rapidly, with each stroke of the driver.

Using a linear analysis, we consider a sine wave at one single frequency at a time.  The impedance (Z) describes the relationship between voltage (V) and current (I), and we can describe the behavior completely if we let V, I, and Z be complex quantities with *real* and *imaginary* parts.  Each quantity can alternatively be described as having a *magnitude* and *phase angle* part.  Both descriptions are useful  depending on the circumstances, and one can convert between them using arithmetic formulas or geometry.  A complex quantity can be represented as a point on a 2D X/Y with x = *real* and y = *imaginary* parts.  Draw a line between this point and (X=0,Y=0).  The magnitude is the length of this line and the phase angle is the counter-clockwise rotation from the positive part of  X-axis to the line.  (This article on the Complex Plane might be helpful.)

For V and I, the magnitude is the absolute value of the peak amplitudes of their oscillations.  The phase angle describes the *phase shift* which essentially describes the temporal shift while recognizing that a continuous sine wave is *periodic*.  Periodic means it precisely repeats at the same interval.  If you shift a sine wave by exactly 1, 2, 3, etc. periods, the result is exactly the same.  So, it makes sense to represent the amount of time shift as a *phase rotation* on a circle.

Analogously to Ohm's law, the relationship at a single frequency is: V = I * Z  or (rearranged)  I = V / Z or (...) Z = V / I.  The rules for multiplication/division of two complex quantities are as follows:  (1) multiply/divide the magnitudes to get the new magnitude.  (2) add/subtract the phase angles to get the new phase.  Therefore the meaning the impedance phase angle is the phase difference or change between V and I.    Note that this math is generally useful for analyzing oscillating signals including audio acoustic transfer functions (i.e. frequency response magnitude/phase).

So if we can easily multiply and divide complex quantities using the magnitude/phase description, what is the purpose of the real/imaginary description?  The latter is for adding subtracting such as when analyzing a series electrical circuit or acoustic interference effects.  The rule for adding and subtracting complex quantities is to add and subtract the real and imaginary parts independently.  For impedance, the *real* part is called the *resistive* part, and the *imaginary* part is called the *reactive* part.  As expected, a pure resistance (i.e. straight a wire) contributes only to the real/resistive part of impedance, and the tendency for oscillating energy storage/release manifests only in the reactive part.  Ideal (as in zero resistance) capacitors and inductors only contribute to the reactive part.  Do note that the resistive part of the impedance is not always equal to the DC coil resistance.  The acoustic properties of the subwoofer contribute to impedance (both resistive and relative) as well.

As such, we can conclude that increasing the DC resistance by heating the coil will alter the frequency response by different amounts depending on both the magnitude and phase of Z.  The increase in Re adds directly to the resistive part of Z.  The peak current is reduced, but the amount of reduction depends on how much the *magnitude* of Z changes. So you can plot the initial value of Z on the Complex Plane and then plot the *new* Z shifted to the right by the increase in Re.  Then, the current will decrease in inverse proportion to the change in distance between each point and (x=0,y=0).

 

[peniku8]

Since I have never worked with the IPAL system before I can only speculate.

Wouldn‘t it be possible to let the amp measure induced current by the driver and calculate inductance from that? Would it then be possible for the amp to actively counter that induced current?

The system could surely have a ambient temperature reference FR and a high temp FR which could then be ‚faded in‘ depending on the state calculated via the dc resistance. Would be some work to set that up but what isn‘t