As it is usually the case, controversies where both (or more) sides look at first sight armed with sound arguments, may carry on forever with no apparent solution. Yet on closer scrutiny some facts surface that show contenders were unknowingly talking different things.
It is basically true the more active components in the amplifying chain, the more the spurious commponents are being added, and this should talk for keeping the line of passive solutions whenever possible.
It is also true that there are thresholds in perception, beyond which no matter the type of artifact introduced by the process, the auditive experience does not change.
Take for example the consumer level 16 bit quantization standard. Purists will argue 20 bit and 24 bit professional level quantization sounds better, and this is certainly true for educated ears. By the same token it could be said 32 bit is much better, doing away with truncation artifacts if you bring it down to 24 or 20 bits. Of course the added 6 or 8 bits in resolution add nothing to the hearing experience for the improvement lies beyond the threshold of perception.
In the case of active crossovers, even inexpensive quad audio quality chips and regular components introduce noise and distortion components beyond most listeners perception limits. In return they provide a level of predictibility and stability far in excess of what an average diyer can obtain when tasked with the design and construction of a passive network that has to be tailored to the individual drivers to boot, just to come near the performance level of the former.
Of course in the end the listening experience including room and environmental issues is the final judge, where the expectations and ideas one carries play a decisive role.
Rodolfo
It is basically true the more active components in the amplifying chain, the more the spurious commponents are being added, and this should talk for keeping the line of passive solutions whenever possible.
It is also true that there are thresholds in perception, beyond which no matter the type of artifact introduced by the process, the auditive experience does not change.
Take for example the consumer level 16 bit quantization standard. Purists will argue 20 bit and 24 bit professional level quantization sounds better, and this is certainly true for educated ears. By the same token it could be said 32 bit is much better, doing away with truncation artifacts if you bring it down to 24 or 20 bits. Of course the added 6 or 8 bits in resolution add nothing to the hearing experience for the improvement lies beyond the threshold of perception.
In the case of active crossovers, even inexpensive quad audio quality chips and regular components introduce noise and distortion components beyond most listeners perception limits. In return they provide a level of predictibility and stability far in excess of what an average diyer can obtain when tasked with the design and construction of a passive network that has to be tailored to the individual drivers to boot, just to come near the performance level of the former.
Of course in the end the listening experience including room and environmental issues is the final judge, where the expectations and ideas one carries play a decisive role.
Rodolfo
(JPK) If we stay in the analog domain it's really not a question of which is better. Active can be good and so can passive. Part of the problem is that many active designs are like the Linkwitz Orion approach; just keep cascading stages until you get what you want, while others don’t address response irregularities at all. More sophisticated active designs are much more like passive designs and use as few active stages as possible to achieve the desired transfer function. The idea that the amp has better control of the speaker isn't really accurate either. If a passive crossover is designed with consideration of the series resistance and its effect of the drivers' Qts, then it's really a wash. With active pole shifting filters and constant current source designs actually remove control of the woofer, for example, from the amp.
Digital is another ball game all together with another set of advantages and disadvantages. The issue of the amp interfacing with the d rivers is still there and must be considered with either active analog, active digital or passive.
Lastly, you might want to consider the apparent contradiction; if active crossovers are in general superior to passive, then wouldn’t active preamps be superior to passive types? So why one hand is there a push to eliminate active line stages in favor of passive ones and then decide to insert an active crossover with a handful or more active stages in there? How can you have it both ways?
As far as I can see it's really just another area where people develop preferences for what they want. At the commercial level it becomes one of cost, as has been addressed in other posts.
Personally I think there is good reason to employ an active crossover between the main system and the woofer (100Hz and down) but between tweeter and mid I see little advantage.
Digital is another ball game all together with another set of advantages and disadvantages. The issue of the amp interfacing with the d rivers is still there and must be considered with either active analog, active digital or passive.
Lastly, you might want to consider the apparent contradiction; if active crossovers are in general superior to passive, then wouldn’t active preamps be superior to passive types? So why one hand is there a push to eliminate active line stages in favor of passive ones and then decide to insert an active crossover with a handful or more active stages in there? How can you have it both ways?
As far as I can see it's really just another area where people develop preferences for what they want. At the commercial level it becomes one of cost, as has been addressed in other posts.
Personally I think there is good reason to employ an active crossover between the main system and the woofer (100Hz and down) but between tweeter and mid I see little advantage.
Active components
Sorry, just for the sake of clarity:
There is no such thing as active capacitor.
Active components are those that convert raw power (usually DC) to signal controled by other signal. Things like discrete transistors, integrated amplifiers or digital gates, and vacuum tubes.
Pushing a little the concept in a usefull way, semiconductors in general are said active though a diode is not really an active device.
Every other component is passive.
Rodolfo
Sorry, just for the sake of clarity:
There is no such thing as active capacitor.
Active components are those that convert raw power (usually DC) to signal controled by other signal. Things like discrete transistors, integrated amplifiers or digital gates, and vacuum tubes.
Pushing a little the concept in a usefull way, semiconductors in general are said active though a diode is not really an active device.
Every other component is passive.
Rodolfo
"With an active crossover, the amplifier is connected directly to the driver, and the only thing between them is the loudspeaker cable. The amplifier presents the maximum damping factor at all times, regardless of frequency, and is not affected by the crossover network, since that is also active, and located before the power amp.
The loudspeaker driver now has the maximum control that the amplifier can provide, across the entire frequency range - not just the crossover network's pass band. The difference in damping is quite obvious, and although some (very well behaved) drivers will show little improvement, the vast majority will be much better controlled, and this will show in an impulse measurement. Not at all uncommonly, it will also show up on a swept sinewave frequency response measurement as well, with the amplitude of peaks and dips generally reduced (albeit marginally in most cases)."
http://sound.westhost.com/biamp-vs-passive.htm
But you're forgetting that real passive components aren't as linear as the ideal. Inductors in particular stray far from the ideal, so do electrolytic capacitors. Furthermore, passive crossovers can cause all sorts of bizzare interactions with the amplifier, causing even more nonlinearities. Plus the increased output required into a passive crossover means the amp will be working harder, which usually also means increased distortion.markp said:
Personally I don't see a reason to prefer a passive preamp to an active one except for simplicity/cost, but in answer to your point: Preamps don't have to deal with large power levels or complex impedances, and can thus be composed of more reliable, closer to ideal passive components rather than large, expensive inductors and capacitors.john k... said:
So what is the story with impedance? Everyone says that impedance is in favour of active designs, as active crossovers dont' need impedance eq like a passive xo. Why is this? Can someone elaborate on the difference here?
So is the fact that the impedance changes throughout a passive crossover a problem for an amp, apart from the fact that at certain points it will be very low?
Is there the need to deal with impedance with an active design? Is there some reason it can be overlooked? Or is it merely the fact that the amp only has to deal with the impedance variation of one driver, rather than two or more at the same time?
So is the fact that the impedance changes throughout a passive crossover a problem for an amp, apart from the fact that at certain points it will be very low?
Is there the need to deal with impedance with an active design? Is there some reason it can be overlooked? Or is it merely the fact that the amp only has to deal with the impedance variation of one driver, rather than two or more at the same time?
some more food for thought ...
some more food for thought ...
If we compare the total amount of distortions inserted by the passive crossover (especially at high power levels), to the total amount of distortions inserted by active crossover, we will find that passive crossover is not as good as active crossover. Even passive components can produce a lot of unwanted effects.
exerpt from crossovers.com
I'm still forming my opinion on this. I was also looking at the website of Newform research. They are making claims that digital crossovers with digital amps are the way to go, that the sound quality is improving, cost is going down and that this will all replace what we now consider as "high end" with a substantial increase in value for money.
http://www.newformresearch.com/
some more food for thought ...
If we compare the total amount of distortions inserted by the passive crossover (especially at high power levels), to the total amount of distortions inserted by active crossover, we will find that passive crossover is not as good as active crossover. Even passive components can produce a lot of unwanted effects.
exerpt from crossovers.com
I'm still forming my opinion on this. I was also looking at the website of Newform research. They are making claims that digital crossovers with digital amps are the way to go, that the sound quality is improving, cost is going down and that this will all replace what we now consider as "high end" with a substantial increase in value for money.
http://www.newformresearch.com/
Look at the basic equations for designing a passive crossover and you will see that they depend on the impedance of the speaker being constant. Now look at the impedance of a real speaker and you will see that it varies wildly with frequency. The result is that unless you compensate for the variable impedance, you end up with a crossover that doesn't work as you designed it.paulspencer said:
A second problem is as you pointed out: The total impedance seen by the amp can drop quite low, which can be hard work for it.
With active designs, the load impedance is of much less importance as the op-amp will correct for any errors. You would be hard pressed to measure any response deviations from an active crossover made with reasonable quality components.
Re: some more food for thought ...
Many capacitors produce "crossover distortion" due to an effect called dielectric absorption. It's essentially a memory effect in the dielectric, similar to the hysteresis in inductor cores. Active filters can use amazing techniques such as "a DC offset" to avoid this effect, passive filters can't. Polypropylene caps have a very low dielectric absorption, and teflon caps are sometimes marginally better (there is some performance overlap) but usually there's no justification for the extra expense of teflon. Passive filters usually need capacitors that have very large values, which can be expensive and can compromise performance just because the components are physically big. I think that the dielectric absorption effect is part of what makes polyester caps sound terrible in a passive crossover, yet they don't sound that bad in an active crossover (with several stages). I'd never resort to bipolar electrolytics
.
Air cored inductors pick up noise from external magnetic fields in the same way that they form their own magnetic field. I'd hate to think what happens when the negative feedback of the amp attempts to cancel out that noise - and there may be lots of it at high frequencies. Thinking about the nasty non-linear dielectric properties of enamel doesn't help when considering the parasitic capacitance of inductors.
Passive crossovers make a mockery of damping factors, and parasitic resistances are just a fraction of the real story. Due to passive crossovers the amp can have very poor control of speakers right across their operating range. This means that break-up modes and box-air resonances will be worse with passive crossovers than without. Note that speakers usually create a lot of spurious voltages by transducing something like 140dB of box pressure and resonances/echoes. The amplifier will attempt to zero-out these voltages, hopefully bringing the speaker's Q down from Qm to Qt, and will thus bring the speaker's unwanted vibrations under much better control (not just at low frequencies). None of this can happen if there's a phase shift caused by a passive component. The spurious voltages that the amplifier receives from the speakers are already delayed, so they can't be cancelled out, and positive feedback that exacerbates the problem is likely too.
Lech
I agree.
Many capacitors produce "crossover distortion" due to an effect called dielectric absorption. It's essentially a memory effect in the dielectric, similar to the hysteresis in inductor cores. Active filters can use amazing techniques such as "a DC offset" to avoid this effect, passive filters can't. Polypropylene caps have a very low dielectric absorption, and teflon caps are sometimes marginally better (there is some performance overlap) but usually there's no justification for the extra expense of teflon. Passive filters usually need capacitors that have very large values, which can be expensive and can compromise performance just because the components are physically big. I think that the dielectric absorption effect is part of what makes polyester caps sound terrible in a passive crossover, yet they don't sound that bad in an active crossover (with several stages). I'd never resort to bipolar electrolytics
.Air cored inductors pick up noise from external magnetic fields in the same way that they form their own magnetic field. I'd hate to think what happens when the negative feedback of the amp attempts to cancel out that noise - and there may be lots of it at high frequencies. Thinking about the nasty non-linear dielectric properties of enamel doesn't help when considering the parasitic capacitance of inductors.
Passive crossovers make a mockery of damping factors, and parasitic resistances are just a fraction of the real story. Due to passive crossovers the amp can have very poor control of speakers right across their operating range. This means that break-up modes and box-air resonances will be worse with passive crossovers than without. Note that speakers usually create a lot of spurious voltages by transducing something like 140dB of box pressure and resonances/echoes. The amplifier will attempt to zero-out these voltages, hopefully bringing the speaker's Q down from Qm to Qt, and will thus bring the speaker's unwanted vibrations under much better control (not just at low frequencies). None of this can happen if there's a phase shift caused by a passive component. The spurious voltages that the amplifier receives from the speakers are already delayed, so they can't be cancelled out, and positive feedback that exacerbates the problem is likely too.
Lech
It was alluded to earlier, but active speakers are not viable in the audio market. There's been many active speakers which many respected audio names have tried to market, and they just don't catch on. For the DIY'er, active is a great way to go, and you can do driver contouring, crossover slopes, and other trickery, which are very hard to achieve (and can never really be done well) passively.
I don't understand why active speakers aren't viable, when the people pay car prices for a line level pre-amp, or thousands of dollars for the last 4 feet of power wire. But audio is in decline (much of it is the industry's own fault), and nobody seems to want an active speaker. Commercial speaker makers not only have to make products that they like, but also make products that will sell.
Sheldon
I don't understand why active speakers aren't viable, when the people pay car prices for a line level pre-amp, or thousands of dollars for the last 4 feet of power wire. But audio is in decline (much of it is the industry's own fault), and nobody seems to want an active speaker. Commercial speaker makers not only have to make products that they like, but also make products that will sell.
Sheldon
paulspencer said:
The problem with passive xo design is that people assume that the network is looking into a resistive 8 ohm load -- the actual impedance, however, is a network of electrical (and "transformed" into electrical components) impedance. But it certainly can be modeled. The advantage of CLC or LCL passive xo's is in their phase response.
Now the problem with an active xo is that the time response or phase delay differs depending upon the topology, how closely you match the resistors and capacitors, characteristics of the operational amplifiers etc. Some active designs are very sensitive to the RC matching.
When I said "gyrators" yesterday I was stirring up the pot -- and I wish more folks would read up on them either in the Burr Brown piece I cited or in the application notes for any number of DSP chips, or the AD713-- a gyrator is "component substitution" which allows you to implement a CLC or LCL xo with only R and C -- you get the advantages of a passive crossover at the expense of a number of specialized operational amplifiers.
And if you think I am kidding about matching R and C components, try building a 1kHz notch filter with randomly chosen 10% components.
In the pro-sound studio monitor market they enjoy a major market share; the consumer market will catch up in due time. What limits their sale ability in the consumer market is the fact that 5.1/7.1 receivers dominate the market. When 5.1/7.1 tuners become widely available the powered speaker will make its way to the consumer market.
And active speakers have already been accepted for computer sound output, most are active now.
It's just the fussy "hi-fi" folks who are reluctant to change.
A review of a speaker with active bass crossover and sub amp read;- "..ultimately limited by the onboard (Mission) electronics..."
Marketing has to counter this negativity first before active systems are accepted. They look at their nice active system with perfect FR etc, and think "... but if I upgraded the amps to Krell...."
It's just the fussy "hi-fi" folks who are reluctant to change.
A review of a speaker with active bass crossover and sub amp read;- "..ultimately limited by the onboard (Mission) electronics..."
Marketing has to counter this negativity first before active systems are accepted. They look at their nice active system with perfect FR etc, and think "... but if I upgraded the amps to Krell...."
From what I understand the fact that a drivers impedance varies with frequency doesn't present any problem at all for the amp with both passive and active designs (as long as your amp has the current available to power that driver at it's lowest impedance). I believe the problem results from the fact that the component values in a passive crossover required to achieve a particular crossover frequency change with respect to the impedance of the drivers. This is why you calculate the values for the "nominal" impedance, it's like an average, but it is actually just a generalization of the drivers impedance curve. Active crossovers don't have this problem because they are line level and are not connected directly to the driver.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.