I can't say for certain, but breaking the speaker load up into segments can surely alter the power transfer characteristics of how each amplifier in a multi-amp system responds. This is in comparison to how one of the same amps driving the entire speaker full-range would respond.
If one factors in that each amp will have differences due to production tolerances, and whether or not each amp is up to the task to begin with, only adds further variables to the equation. This is not to comment on Dazzdax's specific amps and / or installation, but to say that not all amps are created equally regardless of what the paper spec's say and that some simply aren't suited for driving specific types of loads.
Having large multi-driver acoustic suspension speakers with a similar crossover point ( appr 200 Hz ) that i've passively bi-amped, my experiences are probably not that far off from what Dazzdax experienced. The difference here is how we interpreted the changes in presentation and the type of gear being used.
The reduction in distortion that takes place with such an installation can sometimes be quite fooling. Not only does this alter our perceived level of volume, but also of dynamics. By increasing the headroom and reducing the demands on the amplifier, transient response is improved, distortion is reduced and in some cases, the audio spectrum shifts slightly.
The loss in volume is due to the way that our brain and ears process distortion. Distortion is "grating" to our ears, telling our brain "this is loud". Removal of that distortion removes that agitant, resulting in less perceived volume. This is why many folks find themselves playing their system louder than they used to prior to a component change, although it doesn't sound as loud to them at the time. Less distortion equals less "apparent volume". Extreme levels of distortion become more apparent for what they are, telling us "this is not necessarily loud, it is just extremely distorted".
This also factors into both the high frequency dynamics and the shift in the audible spectrum. Due to a reduction in distortion and improved transient response, there is less smearing involved. Smearing is a by-product of non-linear harmonic distortion ( THD ), slewing induced distortion ( SID ) and ringing ( poor transient response ), all of which manifest themselves in a more prominent fashion as frequency rises. If you doubt this, take a look at almost all distortion curves and figures in a Stereophile review. As frequency rises, the distortion by-products do too. The wider the bandwidth of the amp ( faster transient response ), the more level the distortion is across the audible frequency range.*
By removing these aspects of high frequency output that one would normally hear, the midrange and treble region becomes both more subdued and less of an "irritant". Our brain processes that as a lack of dynamics, even though the output is cleaner and ( theoretically ) more dynamic.
In terms of increased bass output, this can be contributed to several factors. The first is that the amp can now concentrate all of the available power that it can develop over a much narrower bandwidth. If the amp was current deficient and / or lacking in power supply reserve for reproduction of longer duration notes, we've now freed up more energy to deal with these problems. The end result as we perceive it would be greater output in the low frequency region than what was there before. This output didn't just show up somehow, it was always there. The deficiencies in the previous installation simply didn't allow it to shine through.
This is one of the biggest problems that i've found with "digital" or "switching" type amps. That is, the bass lacks both intensity and duration. What is there sounds very fast and articulate, but much of that has to do with the lack of duration due to the power supply "pooping out". By reducing the duration of the note, it sounds both "faster" and "sharper" but lacks weight ( duration ) and impact ( intensity ). I've commented on this before as it was something that i ran into with several different types of "new technology" amps, including my Sunfire Sig's.
In most speakers that suffer from poor damping ( aka vented designs ), this can actually sound more natural, even though it is a distortion of the original signal. In this case, one positive error ( excess bloated output from the speaker ) combined with a negative error ( reduced truncuated output from the amp ) sums to a relatively flat presentation. Some would call this "system synergy" whereas others would call it "complimentary colourations". In this specific case, they are basically one and the same.
In Dazzdax's specific case, his speakers don't lack damping ( highly damped low Q sealed design ), so this is not the case. Instead, the increased output capacity of the amp was directly translated into increased output at the speaker. Since this speaker is a relatively linear transducer of energy, more input equalled more output. Combine that with the simultaneous loss of "apparent" mid and high frequency energy and you end up with a completely different presentation.
Given that the electrical characteristics of the speaker remain the same on the whole i.e. separating the top from the bottom doesn't change the nominal impedance or transient capabilities of the speaker itself, the only thing that did change is how the amplifiers themselves would load up.
Bare in mind that this is kind of a "crash course" on this subject. It should help some to understand why they might have run into specific situations when trying to passively bi-amp, but in no means can i hope to cover all the variables involved. Not in one post or a hundred posts. Sean
>
* Negative feedback can be used to try and flatten the distortion curve on an amp, but even then, the distortion will typically rise as frequency climbs. While a completely different subject, slewing induced distortion ( SID ) typically increases as more negative feedback is used due to a reduction in response time of the amp. In plain English, more negative feedback means a slower amp. This can result in increased glare and high frequency stridency i.e. "Solid State sterility". The spec's on paper look good in terms of distortion, but that's only because all of the spec's aren't available to fully interpret the performance of the circuit on the whole.
This is why i've always stressed high speed, wide bandwidth designs in every aspect of the system. This approach naturally combats distortion without the need for "band-aids" that only introduce other problems into the equation.
