IME pure current drive seldom is optimal with resonant drivers, no matter if the resonance is within the passband or not. AMTs worked well with pure current drive, though.
Standard VC-type drivers almost always require some sort of electrical damping around resonance, which leads to a drive impedance profile that is low only around resonance and just low enough to still have resonable system damping (Q < 1 or so).
The point is that a driver will always ring with this Q under any internal error signal or external exitation (adjacent drivers). Optimizing this aspect, well-damped "mechanical" response, IMHO helps the speaker much more than the last bit of measured low distortion squeezed out.
KSTR, in general, you want as little mechanical damping as possible, because all forms of mechanical damping are non-linear and some produce noise. Qms should be high for quality speakers.
Yep, that's why (frequency selective) electrical damping is to be preferred, established in whatever way to make it as linear as possible. The only valid mechanical resistive damping shall come from air load ;-)
EDIT: the true mechanical Qms values we want to compare should be taken in vacuum (no air load) and with non-conductive formers (no eddy current braking). That's the quantity we may want to maximise, not the lower apparent Qms with air load and any nominal conductive (slotted) VC former.
EDIT: the true mechanical Qms values we want to compare should be taken in vacuum (no air load) and with non-conductive formers (no eddy current braking). That's the quantity we may want to maximise, not the lower apparent Qms with air load and any nominal conductive (slotted) VC former.
Some of us know enough about elementary electric circuits to realise that adding a resistor (or anything else) in parallel with a voltage source does not modify the voltage or current in another parallel load. Joe does not realise this, and for reasons known only to yourself you are cheering him on. We do not need to be loudspeaker experts to know this.john curl said:
In such discussions an expert might not add very much, partly because he may assume that nobody could be so ignorant as to assert what Joe is asserting so maybe the person is saying something different. My experience in several different aspects of life is that some people can be slow to appreciate just how profound is the ignorance of some other people and so they defend them.
Yes, but Joe says that it doesn't matter if you use a high resistance cable with one of his speakers because the compensated speaker resistance eliminates the effect of source impedance.RNMarsh said:
My recollection is that flaws in his experimental technique were pointed out. Flaws which were sufficiently serious to largely invalidate his results IIRC.john curl said:
Hi,
Nitpicking, some mechanical damping is reasonable linear. Picture showing Carlsson OA5.
/örjan
Nitpicking, some mechanical damping is reasonable linear. Picture showing Carlsson OA5.
An externally hosted image should be here but it was not working when we last tested it.
/örjan
Not nitpicking, essential for low THD. Look for example at the specs of the Purify midwoofer, the lowest distorting dynamic driver I know.
I have to call this out because it's all too common in this thread to forget lots of "physics" are based on simplifying assumptions.
The original sentence should read: "Some of us know enough about elementary electric circuits to realise that adding a resistor (or anything else) in parallel with a perfect voltage source does not modify the voltage or current in another parallel load..."
Strictly speaking you are right, so it would be interesting to see if adding a resistor (or anything else) in parallel with an output of the imperfect amplifier (like this one, e.g.):
Pass Labs Aleph 3 power amplifier Measurements | Stereophile.com
(or this one):
Ayre Acoustics V-3 power amplifier Measurements | Stereophile.com
would reduce measured acoustical distortion of the speaker. Someone should make such measurements 😉
Could you explain why the mechanical resistive "load" show in the picture would lead to THD?
/örjan
Thank you for noticing. As I repeatedly complained, mere mortals like me has no idea what to expect distortionwise acoustically. Extremely few people seem to care. I noted none of amplifier reviewers does, they are either among the gifted who directly perceive audio directly from electric voltage or ...
Last edited:
It may be small, it may be large, it may be hard to measure, but also this solution could add to distortion. A possible vector is the following
The layer of damping material will move under the pressure of sound energy, it will compact and bounce back. This will cause a non-linearity in the material which will translate into a non-linearity in air resistance. This is all you need to get distortion.
Damping of standing waves = good, damping of cone movement = bad.
Edit: I had to look up this speaker model. I don't think in this particular case it can make matters much worse.
Last edited:
The reason I called it nitpicking is that this specific way of adding restive damping is unusual, I haven't seen it used other than by Carlsson.
Agree on the vector that clearly could be adding distortion, but at least in theory that's an engineering problem to get right. In practice it might be hard to get perfect, as usual in engineering 😉
/örjan
Agree on the vector that clearly could be adding distortion, but at least in theory that's an engineering problem to get right. In practice it might be hard to get perfect, as usual in engineering 😉
/örjan
That's not correct. A shorting ring on the pole piece helps to linearise VC modulating indcutance. It will also help to lower the VC inductance.
If the there is a shorted turn, the speaker is NFG. So if an Aluminium VC former is used, it will always have a slit so as not to become a shorted turm.
T
Hi Terry,
You are correct that the shorting ring help to linearize the VC positional inductance change.
However, the technique uses the conductivity of the ring to form eddy currents which create magnetic fields which are opposing penetration of those fields. As such what it does is exclude magnetic field by opposition.
When the natural magnetic field of a coil is prevented from travelling freely in space using this technique, it reduces the inductance.
The shorting ring excludes magfield therefore lowers inductance. This is used to try and balance the inductance increase caused by the vc going in where more of the vc coils are now included in the magnetic circuit.
As an added component, the shorting ring also dissipates energy due to the eddy currents. That is a velocity dependent dissipation, and dissipates more when the vc is behind the rest position than it does when it is forward of the rest position. The interesting thing, is both flux exclusion and eddy dissipation are absolute value of velocity dependent, not direction dependent.
VC driving out, there is less communication with the magnetic circuit, so the permeability of the flux path is 1, that of free air. That flux is not dissipated, but returns to the coil as the coil pulls back into the gap.
If you look in my gallery, you will see measurements I did and posted of three air core inductors in free air and against a copper printed circuit board plane. I show how the inductance is affected by the eddy exclusion, as well as how the energy dissipation changes as well.
I will try to put the link here, we'll see if I know what I'm doing...
Inductance measurements and eddy effects - My Photo Gallery
inductor_R_characterization - My Photo Gallery
Wow, I'm actually starting to become competent in this posting/linking thing.... Well, at least less incompetent..
The former will also eddy drag, that too absolute velocity and position based, even aluminum that has been split.
jn
You are correct that the shorting ring help to linearize the VC positional inductance change.
However, the technique uses the conductivity of the ring to form eddy currents which create magnetic fields which are opposing penetration of those fields. As such what it does is exclude magnetic field by opposition.
When the natural magnetic field of a coil is prevented from travelling freely in space using this technique, it reduces the inductance.
The shorting ring excludes magfield therefore lowers inductance. This is used to try and balance the inductance increase caused by the vc going in where more of the vc coils are now included in the magnetic circuit.
As an added component, the shorting ring also dissipates energy due to the eddy currents. That is a velocity dependent dissipation, and dissipates more when the vc is behind the rest position than it does when it is forward of the rest position. The interesting thing, is both flux exclusion and eddy dissipation are absolute value of velocity dependent, not direction dependent.
VC driving out, there is less communication with the magnetic circuit, so the permeability of the flux path is 1, that of free air. That flux is not dissipated, but returns to the coil as the coil pulls back into the gap.
If you look in my gallery, you will see measurements I did and posted of three air core inductors in free air and against a copper printed circuit board plane. I show how the inductance is affected by the eddy exclusion, as well as how the energy dissipation changes as well.
I will try to put the link here, we'll see if I know what I'm doing...
Inductance measurements and eddy effects - My Photo Gallery
inductor_R_characterization - My Photo Gallery
Wow, I'm actually starting to become competent in this posting/linking thing.... Well, at least less incompetent..
The former will also eddy drag, that too absolute velocity and position based, even aluminum that has been split.
jn
Last edited:
I suppose if you like fiddling around with those ghastly tube amp things and that open loop, high distortion stuff your comment may apply.
If you use a decent solid state amp, DF96's statement covers it perfectly.
Fair enough.
At least (hopefully) you would go home feeling that your job made a difference and that's not a bad thing. Plenty of people don't.
The Skunk Baxter vid was interesting, I actually haven't seen it before. The musician / improvisational thing explains a lot with my approach to things. I had never really thought about that before.
I'll have to think about that +FB vs NFB issue a bit.
My main point was that the 0.15 'sense' R is very small in relation to a typical say 6 ohm VC resistance (ratio = 40) so, given the 'normal gain; of the amp is not very high, the effect would be very subtle, as far as I can see.
I'll check that article link out later. Not much time ATM.
T
- Status
- Not open for further replies.