I just wanted to throw in a few thoughts here.
This is a topic that has interested me for a while. What are the benefits of biamping, active crossovers, and why for each?
The conclusion I’ve drawn is that for the most part there is very little benefit to either over a well implemented passive design, but it’s still a superior approach for a number of reasons and something speaker designers should use where possible. Consumers should largely not worry about it.
Let’s start with bi-amping. A lot of benefits have been attributed to it such as more headroom and reduced distortion. I don’t believe they either is a real benefit. Let me explain why I feel that way:
-the reduction in amplifier distortion when its highpassed is really small within its linear range. Since amplifiers already have inaudible distortion any reduction in TIM or other non-harmonic distortions is so minimal as to be inaudible.
-I believe that the audible benefits attributed to biamping is actually an increase in dynamic range and that this could just as easily be achieved with a larger amplifier. However, if someone is using a receiver with this capacity and has no ability to add amps, it’s peobably a worthy thing to do. It may add a few dB of extra headroom. It certainly won’t hurt anything. Still, all else being equal, a bigger amplifier makes more sense. The woofers generally draw the most power and as such you may be exceeding your receiver in the woofer alone. The little bit of power the tweeter uses is too inconcequential to matter.
Some may suggest that an active crossover is superior because the amplifier only produces the frequencies it needs to in driving the speaker. That this is somehow not true of a passive crossover. That would actually suggest a misunderstanding of how a passive filter operates. Passive filters do not dissipate energy as heat below the crossover frequency (for a highpass filter). Instead they modify the load of the speaker that the amp sees. This means that the speaker presents a very high impedance load below the crossover and if there is no lowpass filter and woofer, the amplifier would not produce those low frequencies with any intensity. In other words, if you look at the transfer function or a tweeter and it’s highpass filter in isolation and find that it only draws 15 watts to produce 100dB. Then that’s all the amplifier will draw (crossover losses not withstanding). Crossover losses themselves are not that great. The series resistor in an l-pad is the greatest loss to heat and we still see that most speakers rarely could draw more than a few dozen watts across the tweeter before exceeding the tweeters maximum output. The actual losses toninefficiency here are fractions of a decibel. Again, an L-pad does not reduce the output of a tweeter by simplying soaking up excess power and converting it to heat. It raises the resistance of the load and the amplifier simplies produces less power.
For those who don’t believe me, if you put a load in an amplifier and then plug your soundcard mic input to the amplifier output, while the load might see 100 watts, the mic input would see fractions of a watt. Remove the load and nothing changes but the amplifier is no longer producing 100 watts, it’s now only producing a fraction of a watt. This is precisely how a headphone output on many receivers work. A few low wattage resistors create a high impedance load for the headphone out. Mixed with the high impedance of the headphones and the amplifier simply produces a lot less power.
So the take-away is that active or passive crossovers Both cause the amplifier to produce less power in a bi-amp scenario for the high-passed amplifier. This only really matters if the amplifier is being driven into a non-linear State. Which is possible, but then, a larger amp has the same sonic benefit.
See this diagram for an example:
View attachment 27517
So what about the benefits of active crossovers themselves? Well, here I see benefits but it’s not cut and dry for me. First, to go back to the classic argument of DBT’s as the gold standard for anything we do subjectively. I know or no studies where active vs passive were compared under tightly controlled conditions, nor am I aware of any such studies looking at objective measurable benefits either. Instead it seems to be a handful of selective experiments and conjecture. Some of it I think is valid and some I think is not.
Crossover accuracy: theoretically an active crossover can be more precise than a passive one ensuring greater consistency between speakers. While true, it’s very easy to match parts, hold tight tolerances, and ensure such accuracy in passive crossovers. I see this as false accuracy. I see no reason this could make an audible benefit given how accurate passive crossovers are.
Crossover losses: it is true that passive crossovers have losses that create insertion loss. This wouldn’t cause a sound quality difference but it is wasteful. I see this as more a minor but real problem. Good crossovers with low dcr parts have very little loss. Really bad crossovers might show a 1-2dB loss, but good crossovers are fraction of a dB. I don’t lose sleep over this.
Ability to EQ response: ok I hear this a lot. That an active crossover can create a flatter response. Many point to the extremely flat response of many active studio monitors as proof. In fact, this is a real benefit but one that should be taken in context. Passive crossovers can eq as well. Trap filters can be built which apply PEQ that is more than adequate to create a flat response. The issue is that it is cheap and easy to apply many filters in an active system. It is complex and expensive to do the same with passive parts. Still, most good high end speakers have a handful of trap filters in the crossover. So while an active system may have a flatter response, it’s also possible that this is not audible. All else being equal, I prefer the active approach for this reason, but I also recognize that some of this is academic.
Time alignment: digital crossovers can apply time delay to drivers simply, easily, and accurately. Passive crossovers can do this too, but the circuit is complex and hard to calculate. It’s also limited in its range of values, but...given how close the drivers are to being time aligned typically, often this passive approach is adequate. Further, it’s hightly debates if there is any audible benefit to a phase linear system. A good deal of research seems to suggest that phase coherent is good enough and time delay isn’t needed for this. Again, I see this as an academic benefit that I want if I can have it, but I wouslnt lose sleep over it.
So in conclusion my view is that in a perfect world an active system is better, but a passive system is not as bad as people think. All else being equal, no reason not to go active. Getting back to reality, there are other speaker priorities that make a much bigger difference and should take precedence over biamping and active filters.
