Originally Posted by LTD02
"So this 300hz tone bein input creates a 5th order harmonic at 1500hz. In the case of a TD woofer where a lot has been done to reduce flux modulation and linearize inductance, this 5th order harmonic will be low to begin with. There is also no breakup of 10dB or so due to the dustcap in that region. In a driver where there is no inductance control, this harmonic distortion will be as much as 10dB higher to begin with. Then factor in a 10dB increase in frequency due to the breakup. Not only is this a 20dB increase in magnitude of that 5th harmonic, but the dustcap creates a resonance that does not decay quickly. The 5th harmonic tone will be very high in level and continue for a long period of time. This is highly audible."
highly audible is what i would question. with a 4th order low pass at 300hz, spl is down about 55db by the breakup. that is low enough to not be concerned with it.
You're referring to frequency response, John was referring to distortion. The electronic low pass doesn't affect the 5th harmonic distortion product of a 300Hz input. Peaks in the response can amplify distortion products. Audibility becomes a tricky thing to figure as real use involves more than just one sine wave, but the case described would have different distortion behavior. So far as audibility, remember that the higher the harmonic, the more audible the same level of distortion is.
"It's important to recognize what a graph does and does not show. In this case you are only looking at Le vs. VC position. The variance of Le vs. current vs VC position is another 3D variable which inductance control helps with to varying degrees for different methods. A great case to look at for this is near the Fb of a PR reflex design, a bandpass, or horn. I've measured more than a few cases where excessive Le modulation results unexpected distortion increase despite modest excursion vs. rated Xmax."
yep, but that isn't what was mentioned so i didn't address it. :-)~
You answered a comment about linearity of inductance of the B&C mentioned with Le vs. (x) graph. I only pointed out that this was but one variable, and not all inclusive.Edited by Mark Seaton - 9/25/13 at 4:15pm