What 5th Element says is pretty much spot on. I'd particularily like to emphasize the comment about passive XOs creating rollercoaster-like impedance curves. This very often forces the need for a brute force amp instead of a finesse amp and having to suffer the limitations that brute force puts on the sonics.
WRT Pano's comment, he is likely hearing typical active XOs made with mediocre elecronics. It is hard to make a really good sounding active circuit XO, especially as the order climbs.
Not mentioned yet, and often overlooked, is that active makes it much easier to make use of current amps.
And an aside, Earl's comments on active need to be weighed with his belief that amplifiers are almost perfect and i envision his multiamped setup using a stack of HT receivers.
Active subwoofers. So Earl's speakers use both active & passive crossovers. Often this makes a llot of sense.
As in most things, it comes down to execution, and the art of balancing the compromises.
Digital XOs have a lot of promise, but are still immature (which doesn't say what you can get now isn't useful, just that they are going to get way better) and it would be nice to see the sampling rates go up.
dave
WRT Pano's comment, he is likely hearing typical active XOs made with mediocre elecronics. It is hard to make a really good sounding active circuit XO, especially as the order climbs.
Not mentioned yet, and often overlooked, is that active makes it much easier to make use of current amps.
And an aside, Earl's comments on active need to be weighed with his belief that amplifiers are almost perfect and i envision his multiamped setup using a stack of HT receivers.
Active subwoofers. So Earl's speakers use both active & passive crossovers. Often this makes a llot of sense.
As in most things, it comes down to execution, and the art of balancing the compromises.
Digital XOs have a lot of promise, but are still immature (which doesn't say what you can get now isn't useful, just that they are going to get way better) and it would be nice to see the sampling rates go up.
dave
Passive crossovers are entirely unfit for purpose: they are finicky and fickle, drift with age, interact with magnets and with themselves, severely compromise amplifier control over drivers, they botch a ton of issues they are otherwise unable to correct, they screw time behavior, and also dynamic behavior, so please don't get me started. 
This I disagree with, if the system doesn't need anything other then a simple passive crossover, then going active adds in complexity and cost. It will increase the failure rate and possibly increase the amount of expertise required to operate. There is no replacement for a simple passive crossover if the simple passive crossover achieves all of the necessary design goals.
This I agree with but again I point to the first bit where a lot of the time one doesn't need to go active to take the driver specifics into account. This is more relevant in simple two ways with reasonably well behaved drivers, which should describe the vast majority of individually bought stereo pairs of loudspeakers. Can you go active for a small gain or even a widely appreciated gain(and by this I mean Joe public recognises the gain and would prefer to have it)? Yes. It is going to be a good large scale idea as a way of making money, not really (Joe public isn't interested in it if it costs him more and has the potential to be a hassle). Now whether or not this is relevant is another thing, coming from a DIY website where the goal is superior audio, we do not mind going the extra mile. Also if you're selling to a small niche for which superior audio is the no.1 priority then the extra costs/complexity do not matter as much either, but in a global scale economy, active, as of yet, doesn't have a place, because it is more expensive and it does have a greater chance of failing.
This is something however that I see changing as time goes on as DSPs are becoming cheaper and single chip multichannel class D amplifiers with built in DSP capabilities are starting to appear. It is only a matter of time until these become common place and then going active might be more cost effective and will certainly provide better performance. A cheap multichannel class D amp with 4 advanced biquads per channel is certainly going to provide a better crossover then the single cap on the tweeter, woofer run full range variety of mini systems. And if this actually happens then the general public will start to experience better quality sound as the loudspeaker is by far the weakest part of any sound system.
Resonant peaks are handled in exactly the same way if active or passive, all you're doing is reducing the amount of signal reaching the driver. All active has over passive in this regard is that its effects remains constant regardless of how hot the voice coil is.
This isn't necessarily true, Zaph applies a bottomless notch to some of his metal cone designs by using one small capacitor in parallel with the primary inductor. You can also compensate for driver acoustics fairly easily with a passive crossover if you know what you're doing. Passive crossovers only really become expensive if you're interested in using boutique caps/coils, or you're crossing over quite low where the values start to sky rocket. You'd have to have a pretty complex passive crossover for an active one to end up cheaper, power amplifiers cost quite a lot too remember.
Active can do things that passive cannot. If you need these capabilities then it's a no brainer, you go active (Like Linkwitz does). However as stated before, if you do not need these capabilities then it is generally a waste of money going active.
If you're in the position of building all of the electronics yourself then it is quite trivial, generally speaking, to add in another two more power amplifier channels to a design, providing they are low power. Also building a decent active crossover isn't hard nor expensive. Doing these things in an optimum way, if you're buying and not building, will cost you quite a lot for the luxury of going active though. Not to mention you need the know-how to be able to build the things competently in the first place. It is for these reasons that I have not gone the DSP route until recently, times are changing however and as a result, going this route is now far easier then it was a couple of years ago. As Dave has stated DSPs are only just starting to come into their own, give it another 10 years and who knows what will happen.
Which describes in my opinion the number one limitation of passive crossovers. They can only cut, not boost.
Of course you can, this is after-all part of the thread and it's already been somewhat decided that active only makes sense when in the right application. Low frequency crossovers and equalisation is where active crossovers truly shine and prove their worth, such as in a sub. If there is already a power source and a dedicated amplifier present, it is a lot cheaper to implement an active crossover over a passive one, one TL072 + £1 worth of resistors and caps and a simple zener regulated power supply? Yes please.
Passive crossovers are not difficult to implement/design, if you know what you are doing, nor are they more expensive then an active counterpart unless you're doing something you should really be doing with an active crossover in the first place.
One benefit to a passive crossover is rejection of noise. Sure it's a fairly poor advantage as any really well designed piece of electrical equipment should have low levels of hiss and very low amounts of hum, but these things do exist. If you've got a 2 way passive loudspeaker with a woofer dictated nominal sensitivity of 82dB and you're using a 92dB tweeter. In an active system the tweeter might annoy you because you can hear a faint level of hiss. In the passive system however the tweeters level is cut by 10dB and you can no longer hear it. This is a real world problem and DIYers go to a lot of effort in their self built active systems to reduce the levels of hum and hiss to negligible levels. Also the potential for increasing the levels of hum/hiss goes up when you add an active crossover to the system. To some people this is enough reason to stay passive!
I think you're being somewhat optimistic here. The Abbey's are 29" x 15" x 12" and have a 12" driver and the volume they occupy is around 85 litres. When I think bookshelf speaker I think 5-6.5" mid/bass + 1" dome tweeter. These typically occupy about 10-20 litres, so are considerably smaller. Most significant others would balk at the Abbey's. You can hide a ProAc Tablette Anniversary, you cannot easily hide an Abbey under a doily and a potted plant/table lamp.
This I believe is probably the number one reason why the DIY crowd love active crossovers. They are more predictable in their effect because they aren't dependent on the drivers impedance curve and they are easier to tweak. If however you posses the knowledge and capabilities to accurately simulate and design a passive loudspeaker (people such as Zaph and Earl) then you will have equal success (providing that passive will accomplish all of the necessary design goals) with both passive and active. Now which one ends up costing more will depend entirely on the design.
Digital vs Analog
..would be a more useful comparison.
If the drivers of a loudspeaker system are directly connected to amplifiers designed specifically to drive them and the signal has been conditioned and channelized beforehand by a digital front end (DSP), it would be reasonable to assume that the result would produce sound superior in quality to that of other arrangements.
For most, it is a digital world, sans the scratch of vinyl and the hiss of analog tape. This technology, slowly but surely, is finding its way into loudspeaker systems as well. To argue that such an approach is unable to surpass other vintage analog methods is utter nonsense.
Regards,
WHG
..would be a more useful comparison.
If the drivers of a loudspeaker system are directly connected to amplifiers designed specifically to drive them and the signal has been conditioned and channelized beforehand by a digital front end (DSP), it would be reasonable to assume that the result would produce sound superior in quality to that of other arrangements.
For most, it is a digital world, sans the scratch of vinyl and the hiss of analog tape. This technology, slowly but surely, is finding its way into loudspeaker systems as well. To argue that such an approach is unable to surpass other vintage analog methods is utter nonsense.
Regards,
WHG
Which could be true, but what purpose would that be?
On the contrary they are entirely predictable. Do you think Zaph and Earl and every other competent loudspeaker manufacturer/designer go through tens of hundreds of iterations because their simulations don't match up to the real world? Hardly, they most likely go through a maximum of around five designs that are all very similar and then one is chosen after auditioning. They could even possibly do all of the experimenting with a digital box of tricks, then replicate the design passively once they've decided what sounds the best. If you're in the habit of designing lots of loudspeakers this is one situation where having an active DSP setup makes a lot of sense. It saves you time in trying one crossover to the next and it saves you money in needing to have a large parts inventory.
Only electrolytics will do this and even then a good lytic will probably outlast the use of the loudspeaker.
Only inductors will do this in any meaningful way and you can easily get around this by mounting the inductors away from themselves and loudspeaker magnets and in certain orientations.
Placement of coils in crossover networks
It's precisely this lack of control that you are wanting to create otherwise a passive crossover would not work. How else do you think a cap reduces the level of bass delivered to a tweeter? It tends to effectively decouple the tweeter from the amplifiers output as frequency decreases, if it did not do this, it would not work.
If a passive crossover is capable of dealing with an issue then it can deal with it just fine. If however it is unable of dealing with an issue then it cannot deal with it. They do not however tend to create issues simply by their existence.
They do not screw time behaviour, they modify it and in an entirely predictable way, this is also true for the vast majority of active crossovers, be they analogue or digital.
For this to occur the crossover would need to reduce the peak of a transient to something lower then it should ideally be, or perhaps retard the decay of an impulse. Whether or not a passive component would do the first I do not know, do they show non linearities that change with respect to amplitude if used within their capabilities? The latter is certainly true, series resistance can be added to change the Q of a system such that it will alter the group delay, however it is also predictable and can be factored into the design. In reality though voice coil heating probably has far more impact on altering/limiting dynamics.
From all of this is seems that I could be somewhat against active crossovers, this could not be further from the truth and I use them almost religiously in my main system. The loudspeakers in the kitchen and bedrooms get passive crossovers though, why? Because they are better suited.
Yes, like that awful POS the Kaneda active crossover. And those valve jobs. Horrid.
I don't mind the active crossover fanboys, it leaves the passive filters to the real designers.
gedlee said:
Passive xo can easily reach $400-$800 mark. Big inductors are expensive. Then build another speakers, and more investments are needed.
MiniDSP can be had for $200 and you'd get 8 channels. Build another speakers just by uploading new config.
Great amps can be had for $100 -- and that's 8 channels.
Again I see no reasons for passive, except when using single stereo amp is desired. Or perhaps someone who wants to build a once-off kit and be done with it.
You can bi or tri-amp with passive too, you know.
I thought about a mix actually with passive on the mid horn and HF horn, with passive line level X-O between the bass unit and the mids & highs - depending on how capable the GM70 SE turns out to be it might get to do the bottom end and not run full range like the 300B does..
I once saw a 21 component passive 3 way XO that begs to differ.
I like the ease and versatility of an active unit, the expensive part is having an amp on hand for each pair of drivers. If you have that then...
Unpredictable, difficult to design with, suffers from ageing, completely unsuitable for the purpose. IMO, every one of those "features" applies to drivers!
As for crossovers, I went active decades ago and have never looked back. Maybe it's my imagination, but the sum of several small amplifiers and an active crossover seems like way more power than a single big amp and passives.
As for crossovers, I went active decades ago and have never looked back. Maybe it's my imagination, but the sum of several small amplifiers and an active crossover seems like way more power than a single big amp and passives.
Here's how I see it.
For a 2-way speaker, what's simpler than a well designed passive crossover? You can even slope the baffle to avoid the need for a ladder delay network, or use a horn to the same effect.
But who here uses 2-way speakers? Most of us use 3-ways or 4-ways. Yes, if you're alternatively using powered subs I don't know how you can make the claim that your system is "passive". It doesn't matter if your rolloffs are set acoustically or electronically - the definition of active is about the number of amplifiers used. If your subs are being run off the same amplifier as your mains, only then is your system truly passive, else you do see the importance of an active ""crossover"".
Passive 3-ways are generally a mess, especially if they attempt to reproduce the entire range. If it's a passive 3-way, then it really should be intended to cross to a sub, still. There's some perceptual research that suggests we can't "hear" the distortion produced by a cored inductor, but I can definitely tell you how easy it is to measure. Consider how quick certain people are to reject perceptual research that doesn't match their opinion (IE that a speaker with no rear wave information is preferable, despite all the perceptual research shown by Mirage arguing elsewise) and you kind of get to a question of "whether we can hear it or not in an isolated testing condition, is it really worth it to use a cored inductor? Some say sure, "but only for bass" - that's kind of hypocritical IMO. You either hear saturation or you don't, it's not exclusive to which passband your inductor is saturating.
Anyways let's ignore the issue of air-cores vs cores which is an issue in and of itself (kind of like the "debate" of inductors vs opamps for an analog active crossover). How about impedance. If your speaker is a passive 3-way, (and contrary to the claims some in the "all crossovers are evil so don't use em at all) group might make, some VERY valid designs ARE 3-ways) then how likely is it that it's an easy load to drive?. Plus you're inherently limited in sensitivity by the woofer. The woofer can't be more sensitive, and if it's significantly less sensitive then you need to pad down the rest of the system and waste energy. So did you really save money getting a 1000w, 2 ohm stable amp, when you could have gotten a few lesser amps to handle each individual driver? Use two current hungry bass drivers instead of being limited their impedance. Adding amps never hurt anyone except the most frugal.
Here's the nicest thing about an active speaker: programmable baffle step compensation. Some claim that baffle step compensation is part of their inductor design and so they don't call it BSC. But does that mean their system doesn't need BSC? I think that's a dubious claim. If a system sounds best 5 feet from any boundaries, does it need the same transfer function as if I put it 2 ft from the corner for best WAF? Somehow I'm unconvinced.
Then there's crossover slopes. You see people claim they did the same transfer function actively and got the exact same result - that's BESIDES the point!! The nice thing about a quality FIR active crossover is that you're :NOT: stuck to that transfer function and phase behaviour. You just can't beat the narrow vertical lobe of a 300db/octave or higher linear phase filter, passively. For an IIR filter that'd be 50th order - can you imagine the parts count and phase response of such a filter? The results are a vertical response that virtually mirrors the horizontal response, rather than the severe wideband lobing between drivers causing colorations that require a person to place futons and strange diffusion panels on the ceiling.
Finally, there's consistency with respect to transfer function. The fact is, a driver's impedance changes with its heat build up. Some people miss big metal cone breakups not because they had a bad crossover, but because the driver was getting warmed up and the notch filter litterally missed by a hair. That won't happen with an active crossover.
For a 2-way speaker, what's simpler than a well designed passive crossover? You can even slope the baffle to avoid the need for a ladder delay network, or use a horn to the same effect.
But who here uses 2-way speakers? Most of us use 3-ways or 4-ways. Yes, if you're alternatively using powered subs I don't know how you can make the claim that your system is "passive". It doesn't matter if your rolloffs are set acoustically or electronically - the definition of active is about the number of amplifiers used. If your subs are being run off the same amplifier as your mains, only then is your system truly passive, else you do see the importance of an active ""crossover"".
Passive 3-ways are generally a mess, especially if they attempt to reproduce the entire range. If it's a passive 3-way, then it really should be intended to cross to a sub, still. There's some perceptual research that suggests we can't "hear" the distortion produced by a cored inductor, but I can definitely tell you how easy it is to measure. Consider how quick certain people are to reject perceptual research that doesn't match their opinion (IE that a speaker with no rear wave information is preferable, despite all the perceptual research shown by Mirage arguing elsewise) and you kind of get to a question of "whether we can hear it or not in an isolated testing condition, is it really worth it to use a cored inductor? Some say sure, "but only for bass" - that's kind of hypocritical IMO. You either hear saturation or you don't, it's not exclusive to which passband your inductor is saturating.
Anyways let's ignore the issue of air-cores vs cores which is an issue in and of itself (kind of like the "debate" of inductors vs opamps for an analog active crossover). How about impedance. If your speaker is a passive 3-way, (and contrary to the claims some in the "all crossovers are evil so don't use em at all) group might make, some VERY valid designs ARE 3-ways) then how likely is it that it's an easy load to drive?. Plus you're inherently limited in sensitivity by the woofer. The woofer can't be more sensitive, and if it's significantly less sensitive then you need to pad down the rest of the system and waste energy. So did you really save money getting a 1000w, 2 ohm stable amp, when you could have gotten a few lesser amps to handle each individual driver? Use two current hungry bass drivers instead of being limited their impedance. Adding amps never hurt anyone except the most frugal.
Here's the nicest thing about an active speaker: programmable baffle step compensation. Some claim that baffle step compensation is part of their inductor design and so they don't call it BSC. But does that mean their system doesn't need BSC? I think that's a dubious claim. If a system sounds best 5 feet from any boundaries, does it need the same transfer function as if I put it 2 ft from the corner for best WAF? Somehow I'm unconvinced.
Then there's crossover slopes. You see people claim they did the same transfer function actively and got the exact same result - that's BESIDES the point!! The nice thing about a quality FIR active crossover is that you're :NOT: stuck to that transfer function and phase behaviour. You just can't beat the narrow vertical lobe of a 300db/octave or higher linear phase filter, passively. For an IIR filter that'd be 50th order - can you imagine the parts count and phase response of such a filter? The results are a vertical response that virtually mirrors the horizontal response, rather than the severe wideband lobing between drivers causing colorations that require a person to place futons and strange diffusion panels on the ceiling.
Finally, there's consistency with respect to transfer function. The fact is, a driver's impedance changes with its heat build up. Some people miss big metal cone breakups not because they had a bad crossover, but because the driver was getting warmed up and the notch filter litterally missed by a hair. That won't happen with an active crossover.
Last edited:
For $300 you'd get the 96khz variant. And 8 channels.
Good to know, new to me, That is about halfway to a MacMini, i already own PureVinyl, that should give me 24/192 which is what i am really looking for
dave
The newer higher spec models use 96Khz and have twice the outputs are roughly 3x the price of the original 48Khz model. I still think its a good deal at around $300, I've spent more than that on a single driver that hasn't been anything like as good value.
As for the question of passive vs. active, out of the most accurate speakers I've heard there's been no particular bias to passive or active so once again it all comes down to implementation.
Personally I'm a fan of active because of its a problem solving potential. You've got more flexibility in design, a quicker development cycle and features that compliment the active approach well such as linear phase filters that are greater than 6dB/oct, frequency/phase correction, delays, presets and even post processing effects. Those things doesn't necessarily make it superior to passive but its more powerful to the end user. So if we're talking about the basic crossover with no frills then yes they're comparable but as soon as you start talking about features which most digital crossovers have such as PEQ and driver/room correction then I consider active a superior solution. You can add these features to a passive speaker too but if your prepared to do that then why not simply go active anyway.
Its the same story with cost, neither has a real advantage. You can get active products like the miniDSP for $100 or products like the DEQX or Dolby Lake for thousands of dollars. They fundamentally tackle the same problems but with increasing flexibility and fidelity. On the other hand you can build a passive crossover out of low cost components and again you'll most likely spend less than the $100 the miniDSP costs but on the flip side its entirely possible to spend thousands if you start seeking out the audiophile equivalents. With your own passive designs you also need to have a stock of various components to allow for tweaking if your wanting to get the best from any given design and whilst most DIY'ers accrue these over a period of time they're still an investment.
If I'm feeling fancy I add a super tweeter. 3 more crossover parts.
Can't remember ever hearing a 300dB/octave crossover (maybe at a show) but I just heard a very nice speaker that used lots of overlap in it's passive filters. Not my approach, but it sure did sound good.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.