The example I cited was the destruction of a JBL2235 woofer in a B380 enclosure with the BX63 electronic crossover, which has a 20hz 12dB/oct high pass filter.
The 100W Adcom GFA545 when driven into clipping caused the spider to slam into the top plate and eventually destroy the driver.
Substituting a bridged Crest3501, about 1KW on program material, behaved the same way, except it played about 10dB louder before clipping.
Peak limiters work, if adjusted properly, but can make the system sound 'wimpy'.
An amplifier with a hard clip on the input can play much, much, much louder than the same amplifier with a peak limiter.
It's mainly a question of how much subsonic garbage your speaker alignment can handle, for vented systems it may not be much.
The subsonic garbage is generated within the amplifier by clipping, so a filter doesn't do any good.
Changing the RC time constant of the pole in the feedback loop to be higher than the RC pole formed by the power supply filter caps and the load impedance, and then adding back-to-back diodes across the feedback cap seems to take care of the problem.
|
|
