The purpose of the foam is not attenuation.
It is to do a primitive job of being an acoustic lens.
The idea is that by blocking the highest HF from the top and bottom of the ribbon, the shape of the ribbon is essentially "square". So now the radiation pattern at HF (the unblocked area) is more equal in the vertical as compared to the horizontal. Without this, the higher frequencies have nil vertical dispersion. Still not very good vertical dispersion.
The lower freqs pass more or less unmolested through the foam, in theory.
What you are looking for is an acoustic resistance, which unfortunately doesn't really exist in a form that has no effect on freq response...
Try bi-amping with a level control at the input of the amp. Better idea.
_-_-
It is to do a primitive job of being an acoustic lens.
The idea is that by blocking the highest HF from the top and bottom of the ribbon, the shape of the ribbon is essentially "square". So now the radiation pattern at HF (the unblocked area) is more equal in the vertical as compared to the horizontal. Without this, the higher frequencies have nil vertical dispersion. Still not very good vertical dispersion.
The lower freqs pass more or less unmolested through the foam, in theory.
What you are looking for is an acoustic resistance, which unfortunately doesn't really exist in a form that has no effect on freq response...
Try bi-amping with a level control at the input of the amp. Better idea.
_-_-
What is "this is the proper way"??
Putting a resistor in series to increase the Z the amp sees??
_-_-
I think the person posting the request is thinking about changing the frequency response, not the volume. Then the questions make sense.
High impedance source for an AMT? They are pretty purely resistive. I'd be curious to understand the benefits, except again, to alter the response curve to track the impedance instead of the input.
Did I miss a key element?
Erik
High impedance source for an AMT? They are pretty purely resistive. I'd be curious to understand the benefits, except again, to alter the response curve to track the impedance instead of the input.
Did I miss a key element?
Erik
Well, he said "attenuating" and not a frequency response change.
Regardless, a driver that's purely resistive will not have its frequency response altered by inserting a resistor. It will yield a straight attenuation.....just what he's asking for.
Further elaboration from the OP is required in this case.
Dave.
Distortion reduction -- I measured more than 10dB, especially at lower frequencies -- and reduction of higher order distortion components in favor of higher 2nd and 3rd.
The "back-EMF" of an AMT can be quite nonlinear, which seems to be a major cause of distortion. When driven with high impedance, this error term is out of the picture.
Power compression can also quite severe with smaller AMT, with a max continuous power rating of a few watts and rather small time constants. High impedance drive helps to keep the current constant vs temperature, and current is what actually drives a transducer, not voltage.
There is nothing wrong in using resistors and the evidence supports it but that doesn't help if one believes otherwise. We have even special winding techniques, not that it will present an improvement over standard wirewound at the audio frequencies. I reckon a blanket of some sort might be of help.
I ask again, what are you suggesting as a "high impedance source"??
I'd also be curious to know what the presumed distortion reduction mechanism is??
What one hears from various things depends in part on what one can hear, and what one's system can reproduce. Measuring these things and correlating them back to perceptions are a difficult area which has been batted about for years now.
My personal experience is that any time I've had the opportunity to compare a driver run directly from an amplifier and via a resistor or resistive pad, my subjective impression is that it sounds somehow different. And, I don't come to this wanting it to sound different or hoping, or expecting anything.
Also, I have heard the change in the type of power resistor in the output stage of an amplifier "seemingly" (I'll say it this way, ok?) change the character of the amplifier to some degree. For example switching from a good wirewound to a bulk film of the same value.
I'd also be curious to know what the presumed distortion reduction mechanism is??
What one hears from various things depends in part on what one can hear, and what one's system can reproduce. Measuring these things and correlating them back to perceptions are a difficult area which has been batted about for years now.
My personal experience is that any time I've had the opportunity to compare a driver run directly from an amplifier and via a resistor or resistive pad, my subjective impression is that it sounds somehow different. And, I don't come to this wanting it to sound different or hoping, or expecting anything.
Also, I have heard the change in the type of power resistor in the output stage of an amplifier "seemingly" (I'll say it this way, ok?) change the character of the amplifier to some degree. For example switching from a good wirewound to a bulk film of the same value.
A high impedance source is anything which is is significantly higher than the impedance of the driver, say by a factor of 2, at least. Can be a resistor, while an amp set up to give high output resistance is more clever (don't waste power and voltage).
The distortion mechanism is the distorted "back-EMF". The voltage on the driver terminals is the sum of Rdc*I voltage drop + back-EMF voltage. If the back-EMF is distorted the current gets distorted as well because the amp controls the summed voltages. Whereas when the current is injected regardless of terminal voltage (and that's what high impedance drive does accomplish) the distorted back-EMF voltage "runs into the void", does not re-enter the driver. Further, the power compression (rise of Rdc) which can be quite severe with smaller AMTs is eliminated because the injected current is not depending on the Rdc of the driver.
Last edited:
Did you make sure that frequency response was the same, by applying proper EQ? Otherwise the difference will easily have been dominated by just this FR difference which always will happen unless the impedance of the driver is perfectly flat (100% resistive).
Some would appear to deny that "back-EMF" exists or is a problem... but are we talking about the residual energy due to diaphragm motion that does not correspond completely to the signal driving? Aka "distortion"?
In the case of a test signal that excites the motor+diaphragm and stops abruptly, there is clearly "ringing/overshoot", during which time the speaker acts like a generator.
A high DF amp with NFB will see that signal as a difference compared to the input and apply an equal & opposite signal. In theory reducing the "excess'.
In the case where the impedance is high, and due to a series resistor there is less "distortion" signal sent back through the amp. In addition with a speaker that is not perfectly resistive any impedance change represents a frequency response deviation. For example if the Z goes high then less power is transferred than if the match is 1:1... etc. So, one will get a freq response difference.
The same thing happens with a ZFB tube amp, the max power is at the 1:1 impedance match, less everywhere else.
Not sure if the term "injected" fits the situation or not.
I'm entirely unclear how this results in lower distortion in all instances, if any.
_-_-
Yes. When the microphonic voltage (a better term than "back-EMF") is not a linear copy of the driving current then we have nonlinear distortion when is allowed to re-enter the driver.
No, the voice coil *always* acts as a generator, a part of the voltage is proportional to cone velocity.
Yes, but that is the internal feedback mechanism in the driver, the amp knows nothing about it, it just supplies whatever current is drawn while trying to keep the voltage constant. The amount of current is determined by the driver itself from the voltage accross the resisitive part of its VC impedance, and that voltage is the amp output voltage minus the microphonic voltage.
Again, remember v(t) = i(t)*Rdc + BL*dx/dt
(x is excursion)
Most importantly : as long as that is a linear phenomenon, this feedback can be a good thing.
Linear phenomenon, not distortion.
Linear phenomenon.
You are not alone. Most people do not understand how speakers actually work and that there is an optimum drive impedance profile (complex valued, that is, frequency dependant) where various distortion mechanism are reduced best. With a normal amp, the impedance of the voice coil (and any passive filter networks in front of it, if applicable) is determining the amount of the driver internal feedback (a thing most people don't understand), and why should that be the optimum value? It is not.
Last edited:
Though real, back emf is a totally useless and confusing concept. The impedance curve of the driver is the only thing needed to implement a proper passive filter, assuming the amp is a perfect voltage source. If not, this should be accounted for, being resistive or not. Listening the random sonic effects of a non perfect voltage source, or of a non resistive attenuator is imho a total freaky waste of time.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.