so if i have a driver with a weak motor, a heavy cone and a stiff suspension and then it's running at low volume the mechanical parameters don't crop up? i guess the talk of driver "stiction" is just that, talk.
and help me out here so i can clarify my understanding, magnetic resistance? wouldn't we be looking for high magnetic force(Bl) to overcome mechanical resistance
and help me out here so i can clarify my understanding, magnetic resistance? wouldn't we be looking for high magnetic force(Bl) to overcome mechanical resistance
Last edited:
The linear mechanical parameters are always in effect, but not necessarily the nonlinear ones. It depends on the type of nonlinearity. Stiction would be a factor at low levels should I ever see it in reality, but I have not. I have seen a compliance that had a hysteresis issue where at low levels the resonance would jump up in frequency and then at higher levels would drop down, this was an extreme case, as I said before, and have never seen it since. The driver was not even for use in audio, it was an ANC driver.
A high BL does create a large EM resistance, just as you suggest. The higher the BL the less of a factor the mechanical resistance is.
I said that compliance is the major nonlinearity in any driver - that is certainly always the case. Resistance will, of course be nonlinear if the BL is nonlinear, but BL is seldom nonlinear for small amplitudes.
A high BL does create a large EM resistance, just as you suggest. The higher the BL the less of a factor the mechanical resistance is.
I said that compliance is the major nonlinearity in any driver - that is certainly always the case. Resistance will, of course be nonlinear if the BL is nonlinear, but BL is seldom nonlinear for small amplitudes.
i still can't understand how any part of a loudspeaker's mechanical and electrical operation can in any way be considered "linear".
so what are you referring to when you say "linear mechanical parameters are always in effect"
sorry i thought Bl represented force quite the opposite of EM resistance (as in high Bl - low EM resistance? no?)
so what are you referring to when you say "linear mechanical parameters are always in effect"
sorry i thought Bl represented force quite the opposite of EM resistance (as in high Bl - low EM resistance? no?)
ok thanks for the clarification of your statement sorry i asked, as for modeling there does seem to be some problems i've encountered and do need to understand better, i just thought discussing with, and asking a learned person such as yourself would help the process.
and yes i can appear obstinate and slow to comprehend, no need to border on a personal slight.
and yes i can appear obstinate and slow to comprehend, no need to border on a personal slight.
Last edited:
Since you are so nice about it, let me tell you where you will have trouble understanding. Linear does NOT mean a straight line of frequency response - its nothing that simple. It has to do with how the system reacts differently at high amplitude versus low ones.
In a transducer model the hardest thing to grasp is how the mechanical and electrical aspects couple together. They are intertwined in a highly complex way. The roles of things reverse on each side of the voice coil where large resistances in series become small resistances in parallel. This is not only complex to understand but completely non-intuitive, which is why your intuition is leading you astray.
Look up a Gyrator and try to get your arms around that. The electrical domain is on one side and the mechanical is on the other. Please note that in electric circuits such a thing as a gyrator does not exist - it can be built up, but it takes many components. It exists only as a mathematical nicety to a complex set of relationships.
In a transducer model the hardest thing to grasp is how the mechanical and electrical aspects couple together. They are intertwined in a highly complex way. The roles of things reverse on each side of the voice coil where large resistances in series become small resistances in parallel. This is not only complex to understand but completely non-intuitive, which is why your intuition is leading you astray.
Look up a Gyrator and try to get your arms around that. The electrical domain is on one side and the mechanical is on the other. Please note that in electric circuits such a thing as a gyrator does not exist - it can be built up, but it takes many components. It exists only as a mathematical nicety to a complex set of relationships.
say what?
if a system reacts differently at high amplitude vs low amplitude how is that linear?
and no i don't think linear is a hard straight line, if the amplitude response is to be considered linear it should not vary with volume or at least not deviate more than 3 db from it's nominal state over a wide range of volume(i've yet to find a driver, compression or any other type that does this and no that's not a perception that's from measurement data)
as to gyrator's, there the hypothetical lossless math model which may be mind numbing but i'm more at home with gyrator circuits as used in the Klark DN27a eq to replace large inductors...or am i the victim of Klark Teknik's marketing department and these things don't exist?
sorry if this coming across as rude but it when you say "in electric circuits such a thing as a gyrator does not exist-"(and subsequently say)" it can be built up but it takes many components" is rather contradictory to me.
this article on gyrator circuits suggests to me that as little as a couple of op amps and a few capacitors (low parts count) can produce a gyrator
https://inst.eecs.berkeley.edu/~ee100/fa04/lab/lab10/EE100_Gyrator_Guide.pdf
and i've no illusions as to the complexity of the of the intertwining of mechanical and electrical aspects of loudspeakers which is why i constantly question conventional wisdom.
if a system reacts differently at high amplitude vs low amplitude how is that linear?
and no i don't think linear is a hard straight line, if the amplitude response is to be considered linear it should not vary with volume or at least not deviate more than 3 db from it's nominal state over a wide range of volume(i've yet to find a driver, compression or any other type that does this and no that's not a perception that's from measurement data)
as to gyrator's, there the hypothetical lossless math model which may be mind numbing but i'm more at home with gyrator circuits as used in the Klark DN27a eq to replace large inductors...or am i the victim of Klark Teknik's marketing department and these things don't exist?
sorry if this coming across as rude but it when you say "in electric circuits such a thing as a gyrator does not exist-"(and subsequently say)" it can be built up but it takes many components" is rather contradictory to me.
this article on gyrator circuits suggests to me that as little as a couple of op amps and a few capacitors (low parts count) can produce a gyrator
https://inst.eecs.berkeley.edu/~ee100/fa04/lab/lab10/EE100_Gyrator_Guide.pdf
and i've no illusions as to the complexity of the of the intertwining of mechanical and electrical aspects of loudspeakers which is why i constantly question conventional wisdom.
Guessing here but shure you not run into microphone acoustic clipping area into these tests, not all microphones have printed data for where we have max SPL/distortion limit/clipping numbers going on.
much of the data i observed was years ago, and that was with an RTA i am currently attempting to see if what i noticed in the past i can see again, using something like REW and indeed a calibrated microphone.
i've also been actively searching on the net and through the university library for any similar data or research as to the variability of frequency response with volume/loudness there seems to be a void/absence of info with respect to this.
i've also been actively searching on the net and through the university library for any similar data or research as to the variability of frequency response with volume/loudness there seems to be a void/absence of info with respect to this.
For own diy speaker and hobby systems have not observed any frq response deviations at various levels using REW but also have been relaxed about it and not shall we say zoomed in or did any hunt down that road, on the other hand guess subjective its important dial in a reference level for each and every trial session because else we are misinformed because of our hearing system and the equal loudness curves in ISO226-2003 paper, for that enjoy using JRiver as player because it has features to be level calibrated and correct for those hearing curves if one is okay using digital level inside a 64bit engine.
Frequency response changes with level are very well known and documented. The vast majority of the changes come from thermal changes in the drivers. I have noted this many times here at DIY.
But we were talking about nonlinearities at low signal levels and these thermal changes do not occur there. It should be noted that thermal changes are not rightly considered "nonlinearities" as they do not happen on the signal, but on the system.
But we were talking about nonlinearities at low signal levels and these thermal changes do not occur there. It should be noted that thermal changes are not rightly considered "nonlinearities" as they do not happen on the signal, but on the system.
if as you say "response changes with level are well know and documented" can you please supply me with either links or reading references i'm very interested in any material on this topic.
power compression is well know to me coming from a sound reinforcement background but the response changes i've seen where occurring long before this region of operation.
power compression is well know to me coming from a sound reinforcement background but the response changes i've seen where occurring long before this region of operation.
Of course, as you say, electrical damping = (BL)^2/(Re*Mms) = 2*Pi*(Fs/Qes) is always much more than mechanical damping = Rms/Mms = 2*Pi*(Fs/Qms).
BUT even so, all else being equal and for the same amount of total damping (Fs/Qts = Fs/Qes+Fs/Qms), wouldn't you always prefer a driver for which the share of total damping provided by the electric "motor" (i.e., (BL)^2/Re) is maximised and the share of damping provided by the much less ideal and more compromised mechanical resistance (Rms) is minimized?
That was the point I was trying to make earlier on.
the good old 80% rule: an 80% solution implemented expediently in the present beats a 100% solution never implemented or slowly implemented. it's not perfect, but it's done & it works!
So true, try to explain this to 'low mass' planar guys - in the end I usually just give up.
I wouldn't qualify the JBL 2235 (H) as a movie theater driver. It was only used in the B380 (Sub) and the 4344, 4355, and 4430 Studio Monitors.
The 2235 is a (sub)woofer for studio and domestic use, whereas the 416 is an oldskool PA midwoofer.
An original 2235 features a mass-ring, which affects the performance towards and above 500Hz.
It wouldn't surprise me if you preferred the sound of the 416, even with the polyurethane surround of the 2235 in perfect condition.
Many 2235 owners remove the mass-ring, which improves the midrange and turns it into a 2234.
Last edited: