98.247.165.94
| '); } else { document.writeln(''); } } else { document.writeln(''); } } else { document.writeln(''); } } // End --> |
In Reply to: RE: Loss of "reality" when the volume is turned down posted by Peter Gunn on November 06, 2009 at 10:06:51
PG,Some things are true, whether we believe them or not. I'm afraid your computer is in error....you're misinterpreting the results....your test setup is incorrect.....or some combination of those.
Passive components between the amplifier and transducers can (at best) pass the signal through with no attenuation.
So, the explanation for what you're hearing (which was in my previous posting....please read it) is that you've changed the crossover alignment to "blend" the drivers together much more than a stock crossover. That yields a boosted response relative to the stock alignment but ONLY in the crossover region, not for the full bandwidth of the system. The stock crossover networks yield a fairly flat response over the speakers full bandwidth, but the crossover you recommend does not exhibit a flat acoustic response. That is equalization, not a sensitivity increase.Regarding your quoted number of 92db: Let's take Magnepan's quoted MMG sensitivity as an example. It's 86db for a 2 watt input at 500hz. Since you don't specify, I'm assuming you believe your 92db spl output will result from the same test conditions? That tells me that you're claiming a 6db increase in sensitivity?
Theoretically, that's possible, but only if you coupled together two, identical, full-range transducers, wired them in parallel and drove them full range. Clearly, the MMG's don't satisfy any of those so, by definition, a 6db increase is simply not possible. The actual measurements for the crossover you recommend show a peak relative to the stock alignment of about 3db centered at the crossover frequency, but no change in output level outside of the crossover region. This would be expected based on the simulations and I've confirmed it myself with an actual measurement using the series network you recommend.Now, if your 92db claim is based on an alternative/incorrect reference then maybe it's possible. :)
The real sensitivity number for stock MMG's (based on my own measurements) is approximately 82.5db at 1 meter with 1 watt input averaged over the full frequency range in free-field conditions. This is on par with Magnepan's quoted specifications.
The baseline efficiency of each of the Magnepan models is "built-in" to the transducer design. You can't alter the crossover network to increase that efficiency by 6db (or even 1db.) It just doesn't work that way.
Dave.
Edits: 11/07/09Follow Ups:
Post a Followup: