jay mitchell
Roadie
- Messages
- 620
Curiosity got the better of me, and I just did some tests. I did free-air impedance measurements on an Eminence Legend 1258 I had lying around at four different power levels: approximately 1W, 2W, 4W, and 8W. In a free air test, it takes relatively little power to make a typical guitar transducer go audibly nonlinear in the region of its resonant frequency. Trust me on this: 8 watts is more than enough.
There was no measurable shift in the resonant frequency at any power level, even though the speaker was sounding pretty zippy at 8 watts. FYI, that much power into this Legend in an enclosure would yield about 109dB at a 1 meter distance.
The one thing that changed due to increasing power level was the magnitude of the peak impedance. It went from ca. 55 ohms at one watt to about 53 ohms at 8 watts. The test is most sensitive to environmental noise right at the speaker's resonant frequency, so the difference lies just barely outside test-to-test repeatability. The lumped-element parameter that would cause this kind of variation is the speaker suspension's mechanical resistance, which acts proportional to cone velocity. Given the easily-heard air noise, it appears most likely that turbulence due to air pumping in the magnet gap is the cause of the magnitude shift. I could verify this by removing the dust cover and running the tests again, but I'm not really doing R&D here. I'm posting to make a simple point supported by data I have acquired.
I saved the graphs on my test system's host computer, and it will be something of a project to transfer the files to another box (long story). I'll try to do that later and upload the graphs here.
There was no measurable shift in the resonant frequency at any power level, even though the speaker was sounding pretty zippy at 8 watts. FYI, that much power into this Legend in an enclosure would yield about 109dB at a 1 meter distance.
The one thing that changed due to increasing power level was the magnitude of the peak impedance. It went from ca. 55 ohms at one watt to about 53 ohms at 8 watts. The test is most sensitive to environmental noise right at the speaker's resonant frequency, so the difference lies just barely outside test-to-test repeatability. The lumped-element parameter that would cause this kind of variation is the speaker suspension's mechanical resistance, which acts proportional to cone velocity. Given the easily-heard air noise, it appears most likely that turbulence due to air pumping in the magnet gap is the cause of the magnitude shift. I could verify this by removing the dust cover and running the tests again, but I'm not really doing R&D here. I'm posting to make a simple point supported by data I have acquired.
I saved the graphs on my test system's host computer, and it will be something of a project to transfer the files to another box (long story). I'll try to do that later and upload the graphs here.
Last edited:
