Whaaa? I hope you were not responding to my post, I was just trying to make the point that people are mispeaking when they claim accuracy to a sound they have never heard. I also don't think that the sonic characteristics one person prefers is the exact same as another. If this was the case there would be one mastering suite design and a standardized engineer, and one manufacturer of speakers. We do need the continual advancement in both electronics as you have done and especially in transducers and awareness of room acoustics which, despite it's importance in far-field listening as most people do has taken a back seat because you can't sell it in a store.
I have heavily invested in accurate and well-made electronics to this end so I am a believer in well-designed and executed equipment. The fact that you show adherence to RIAA within fractions of a dB shows how stable and well-quantified electronic measurement is, and as such there is no reason to not design equipment to the best possible measured performance. The process of measuring speakers is far less robust, indeed it involves an additional transducer with it's own characteristics, nevertheless it is extremely valuable at quantifying performance. However, the RIAA equalizer you measure will remain stable no matter whose house it is installed in...but that is not the case with the speakers.
Due to my space and acoustical constraints, for years I have done what I have also heard you espouse: I listen near-field to a set of Tannoy dual-concentrics which image wonderfully...in the nearfield...six feet back, not so much. As it occurs when one uses a set of speakers for a long period, I am normalized to them and can hear relatively small differences in sources and up-stream equipment as a result, so I look forward to incremental improvements in electronics. I also listen to headphones and IEMs which give a different, but often enjoyable experience.
Sorry if my intent was misconstrued!!
Howie
🙂 😎
Oh ! Thx so much for explaining in more detail re. accuracy. Good to be in company of people who know what they are doing.
I will also repeat - go back to line # 2350 from time to time as others come in and out of this forum .
THx-RM
Last edited:
Hi Chris
If only it was that simple. I don't disagree with you entirely, but even with active there is a problem with this idea of 'control' when you have a voltage source driving a current device. The idea that the voltage to current conversion takes place across the voice coil in the way that it can only happen with a resistor, this simply is not happening, just look at the current phase angle (power factor) and you know this cannot be happening. All I know is that at least in some places this is now being looked at. And it applies to both active and passive speaker systems, but clearly a crossover is an added complication.
I do not disbelieve in DSP, but I agree, such things should be kept to a minimum. But to me, that is a different subject. This is about a subject (in my mind) where we really have not yet come to grips with. The driver needs current, yet for the most part, we feed it volts. Yet the one thing a voltage source/amp cannot do control is the current. And yet we persist with a voltage model that is very powerful, yet does not give us the full picture. But I am hopeful that a better picture is now starting to come forth, there does now seem to be some appetite to look into this - and not the fall-back solution that some purists will have, meaning preaching current drive - that is just not a real world solution either. What seems to be emerging is the idea of equalising the current of the amplifier and work on this is being done in the Netherlands (where some speaker manufacturers are doing it behind the scenes) and here in Australia - several physicists and designers/experimentalists are involved. Things like this always starts small.
- Joe
Neither voltage or current are the right answer if you want to directly control the motion of the diaphragm. The control the force at the voice coil and not perfectly. The original Rice Kellog design understood this and counts on less motion for a given voltage at the voice coil to get constant spl.
You can sense the coil or even the diaphragm but its still not ideal. The acoustic wave has parted company with the voice coil such that you can only guess what the diaphragm is doing. Really stiff light diaphragm help BUT have really high when the break up.
Electrostatics can be a better option but the acoustic wave front is still a challenge. Still the current waveform on electrostatic seems to better mirror the diaphragm motion.
You can sense the coil or even the diaphragm but its still not ideal. The acoustic wave has parted company with the voice coil such that you can only guess what the diaphragm is doing. Really stiff light diaphragm help BUT have really high when the break up.
Electrostatics can be a better option but the acoustic wave front is still a challenge. Still the current waveform on electrostatic seems to better mirror the diaphragm motion.
With, say, DC PM servo motors running open loop, the usual tendency is for voltage to roughly control velocity and current to control torque. What is the amplifier output signal supposed to represent?
Anyway, the idea is that the applied voltage mostly sets velocity and the motor should draw current as needed roughly in proportion to the load.
Anyway, the idea is that the applied voltage mostly sets velocity and the motor should draw current as needed roughly in proportion to the load.
Last edited:
Well, directly connected to the amplifier output the entire loop has lower impedance. The back emf from the driver is damped effectively by the amplifier output stage. We'll assume a well designed amplifier to limit the range of responses to the subject at hand.
All we can do is provide enough current capacity so that driver can be controlled. A low source impedance does this. As for what the diaphragm has already done, that is something the driver manufacturer has determined. We can only do so much. I think we can agree that once you introduce an impedance between the driver and source, you lose control over what the driver is doing. This becomes worse as that impedance rises.
-Chris
All we can do is provide enough current capacity so that driver can be controlled. A low source impedance does this. As for what the diaphragm has already done, that is something the driver manufacturer has determined. We can only do so much. I think we can agree that once you introduce an impedance between the driver and source, you lose control over what the driver is doing. This becomes worse as that impedance rises.
-Chris
I hope you will enjoy them !
Dutch & Dutch | The Ear
Regarding the mid frequency directionality passive control:
DIRECTIONAL LOUDSPEAKER
When you will receive the D&D 8c, you will have the opportunity to test this. Place them where your omnis are and please inform us of what you think of the outcome of the [omnis/directional] comparison.
PS.Which omnis do you have?
George
Attachments
Hi Chris
Good to engage with you again - this is actually a very interesting topic, but I cannot help note that with JC's presence, this forum seems to gravitate more towards amplifier design than speakers.
The back emf from the driver is damped effectively by the amplifier output stage? Not necessarily.
Indeed the back-EMF is a voltage source (not a current generator) that opposes the current of the amplifier. The myth of the amplifier adding damping was disputed way back in 1975 when I had that discussion with Richard Small, in those days lecturing at Sydney University.
He was referring to the equation below - and according to it, the amplifier can only make damping worse:
An externally hosted image should be here but it was not working when we last tested it.
Analyse the equation and the Re is the DC resistance of the voice coil plus any additional series impedance, such as the output impedance of the amplifier, resistance in coils and wiring. So unless the amplifier has a negative impedance (making Re less), the amplifier can only erode Qes, not improve it. In the same way that those other things also erode the electrical damping - not much to get confused about here.
There is a lot of misinformed nonsense about this subject - most amplifier designers have promoted this, speaker designers just allow for the amplifier to have a range of output impedances, because they know some users will be using tube amplifiers, so speaker designers rarely think in other terms - the truth is that the alignment is what determines the damping - yes, amplifiers can affect the alignment, but the alignment can be adjusted so that damping is not lost, lower Fb and/or enlarge Vb. So these things are in the purview of the speaker designer, not so much the amplifier designer.
Joe
Last edited:
Hi Joe,
Finally, you're right. There isn't much that we can do about it beyond selecting another speaker model.
-Chris
True and unavoidable since the amplifier is the source of excitation. The only thing we can choose are amplifiers that have a low and linear output impedance. However, the equation really does a great job of pointing out the horror that is a crossover. If you work out the instantaneous impedance of the crossover and add that to the dynamic output impedance of the amplifier, you will see that the crossover swamps out the effects of an amplifier. That's assuming a proper crossover that rolls off the highs from the woofer.
Finally, you're right. There isn't much that we can do about it beyond selecting another speaker model.
-Chris
Chris, if you equalise the current of the amplifier and end up with both voltage square waves and current square waves across the audio band, you can actually cancel out the effect of the series impedance of the amplifier. This is not theory, it has actually been done. The point is to cause the amplifier to supply the same current at all frequencies - and it can be done if extreme care (and design) is used. You can even current drive such a loudspeaker and the magnitude of the response will not alter, any source impedance can be used. Think about it and you will understand why the above is true. 😉
If the load is not resistive you can't have voltage square waves and current square waves at the same time.
What if the amplifier has complex conjugate output impedance?
I don't know, haven't thought about much , but if not that then something more or less like that would have to be going on.
I don't know, haven't thought about much , but if not that then something more or less like that would have to be going on.
Last edited:
It might be difficult to apply to audio, conjugate matching is usually a narrow band RF concept audio is 10 octaves. After all the cages rattle here when treating cables as transmission lines is even mentioned.
Magepans are basically resistive. . . Most transducers aren't. Are you suggesting a series parallel network to counteract the impedance variations? Maybe some nonlinear network in the feedback that reflects the driver nonlinearity?
Definitely parallel - and yes, ribbons and Maggies are basically resistive and I like. 🙂
- Status
- Not open for further replies.