Hi Brian!
> I think once an amplifier is connected to that
> system, (or the voice coil is shorted out) its
> response will be primarily determined by Qes, not
> Qms.
I think I didn't explain myself well, what I was trying to say is motors with real high Qms tend to sound really "muddy". Thats not to say all motors with high Qms sound bad though.
These sorts of "bad" motors are ones that have Qms that is way high, in attempt to tune the woofer very low. This is done to encourage maximum bass extension and motor strength is necessarily weakest. And the magnet is likely to be made of what is probably a cheap non-homogenous ceramic "cake" and an ineffective and non-symmetrical magnet geometry. It is overall just a poor performer, and sounds real bad. Often though, they'll have very low resistance and the value of Qes is "shunted" across the mechanical Qms, which is much higher. Qts is then made relatively low by virtue of the low Qes shunted across what is a fairly high Qms. In this case it needs motor braking in the extreme. But with motors with very high Qms, say 20, they are just real bad, and could be desribed as a "mudmotor". Even the example above, is not a driver I would like - and it has poor filter characteristics, the way the driver has been designed too, is poor, and is not a good performer. Distortion is in the triple digits on a device like this.
> I think once an amplifier is connected to that
> system, (or the voice coil is shorted out) its
> response will be primarily determined by Qes, not
> Qms.
Again, thats not exactly what I was arguing, but being connected to the system - the motors filter characteristics dont change, and these high Qms mudmotors still perform badly.
> Now, short the terminals with a piece of cable and
> tap on the cone again. The boingy sound should
> disappear immediately and the speaker will cease to
> ring (unless its Qes is also high, of course). Qes
> now dominates.
Thats right. Shorting the terminals is like having an amplifier with infinite damping factor, and it should control the cone. But thats not to say Qms is not relevant, because it is - what we have in this situation with the shorted terminals, is what would look like extremely low output impedance of the amplifier - but of course, theres no amplifier there, its just a wire shorting the terminals which also has very low resistance. Output impedance is inversely proportional to damping factor, and it highly affects Qes. Shorting the terminals essentially provides infinite motor braking to the driver. Increasing magnet strength, by more magnet, or more coil, also affects Qes. Of course though, it needs to be connected to the system for it to operate! But you've made a good point though, and motor braking is typical of drivers having high Q and low resonant. And these sorts of "mudmotors" typically have distortion over 100 percent. But we cannot say that Qms is not relevant, because it describes a large part of how the driver is damped and what we can expect of these kinds of drivers, often showing how manufactures cut costs in attempt to maximize bass extension - but throw away just about everything else, to put it loosely.
> Perhaps, but if JBL is referring to a weak magnetic
> structure that's not exerting sufficient control
> over the cone, this suggests that it is Qes is the
> issue, not Qms - the weaker the motor, the higher
> the Qes of the driver. Obviously though, the higher
> the Qms, the greater the problem as Qes itself
> increases (Qts will approach Qms).
Total Q and the figures that make it up, Qes and Qms, are not the only factors of distortion. Having a perfectly symmetrical magnet field, underhung voice coils, voice coil windings, magnet material, flow of magnetic flux, magnetic geometries, copper shorting rings, cooling techniques, and more, all are contributors to distortion, and all primarily arise from the magnetic structure.
> I'm not surprised - excursion rises dramatically as
> frequency decreases. The rate of increase will
> depend on the driver's Qts (the higher the Qts, the
> greater the increase).
The increase in distortion is not just due to an increase in excursion, rather, the entire magnetic structure is responsible for this. This area is what I call the "critical resonance region", and distortion rises as a function of frequency here. You'll also notice that at the upper cutoff of the driver where response starts to drop again - distortion rises too, partly due to breakup modes. But the distortion at resonance is a much bigger concern. All drivers exhibit least distortion midband.
Adrian
|
|