Active vrs passive

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
CopperTop, I've never heard anyone say that active is only for newbies. Maybe back in the day, the only options on the market were analog active crossovers, which had no baffle step compensation option and used terrible sounding op-amps. Even today, the options on the market for digital crossovers aren't great (I've compared, and this is my personal conclusion), though far better than in the past. The DIY options for great digital crossovers is already here today, just expensive.
 
But there is a meme going about that while active are OK for beginners and philistines, an active is no match for a well-designed passive crossover.

This is nonsense.

The argument may be along the lines that passive crossover design is "hard" and the designer has to serve such a long apprenticeship that he inevitably does it better than the idiot with the computer.

Neither passive nor active are any harder to a seasoned designer, this might be missing the point a little, but a skilled designer, who is going to be able to create a work of art with a passive xover, is going to be able to do the same with the active version. Mr Idiot with his computer and active xovers will never reach the same level of perfection as the designer who knows exactly what he's doing and how to go about doing it.

But... we also hear the claim that passive crossover design is all about science and
measurements and objectivity.

Indeed.

Well a computer can calculate in a second what even a member of The Crossover Designer's Guild would take a lifetime to do. There is no magic we're told, so the calculations and formulae are already defined. So let's stick them in a computer and come up with the truly optimum passive crossover and marvel at its beauty.

The trouble is that computers are horrible at noticing patterns, face recognition for example is something that computers struggle at immensely. A skilled designer is often capable of looking at the data provided and know within a minute or two, whether or not two drivers will work well together and then know pretty much what type of xover is going to be necessary to get the job done. The computer helps him fine tune the design saving an immense amount of time, but it doesn't pick the filters.

But passive electronics is all about conflicting requirements and compromises (an 8th order passive is practically impossible, we're told, for example), so the digital active has simply got to be better in strictly objective terms.

If an 8th order active is required then yes things do start to add up. But the first part about conflicting requirements in passive xovers, this is absolutely true, because you do have losses in passive xovers and the simpler they can be made, whilst still doing an adequate job, the better. Often its this balancing act of getting the most bang for your buck of where the human shines over the computer. The human has the ability to again recognise where doing something for simplicities sake isn't going to cut it this time, but why another time it will.

Personally I'm not interested in the slightest in any non-DIY considerations like hardware cost and size, so I don't think they have any place in this discussion; I want to build the best possible system and I haven't yet heard a real, non-magical, argument in favour of passive crossovers.

You wont get one.


At very high level, active analog is not there yet.

This is nonsense.

That might be me. Again, very reasonable, except the passive crossover designers want us to believe that it's so incredibly difficult, what with their having to juggle hundreds of variables written down on vellum (or whatever sounds best...), that the process of condensing all this down into three components takes hundreds of hours.

I think this is exaggerating things a little bit, but being able to reduce a crossovers component count whilst allowing it to keep doing pretty much the same job as it was doing 3 components more ago IS something that design experience gets you.
 
Last edited:
Cool. The audio hobby is crazy. Just mention your speakers, amplifiers, source code etc, and we will see.

Well the speakers are Mission M702e (being taken to to the council dump but I said I'd have them). I fixed the rattle, but they didn't sound great; woolly and muddled I thought, with unnaturally 'warm' bass. And yet... sometimes more pleasant to listen to than my Tannoys with certain types of music.

So I wanted to try active speakers, and these seemed ideal as a basis for experiment. I bypassed the passive crossovers and inserted a 100uF non-polarised electrolytic in series with the tweeters for protection (about £1 each). I bought two secondhand amps: Denon PMA355UK (about £30 each), an old Dell desktop PC which is almost silent (£40) and a Creative X-Fi card (£30). I then spent many, many happy hours writing the software to allow me to play any source in the PC through my own DSP code and thence to the analogue outputs of the card (cost £0). The crossover filters are strictly linear phase and complementary so if you add their outputs together the result is the original signal. If I want, I can obviously apply any sort of further processing to each driver and, if things had not worked quite so well, that's what I would have done. However, when I set the woofer and tweeter levels to be the same at the crossover point using a test tone, and ran the actual system for the first time with real music it was a eureka moment; I'd never knowingly listened to active speakers before, and in an instant I understood what the fuss was all about. Since then I've refined the crossover shapes and slopes, and added a fairly usable graphic interface, but very little else.

Now I don't yet know how lucky I've been in picking the perfect speaker for this conversion (in every sense!) enabling me to get away without applying any further correction, but my three-way Tannoys are about to get the same treatment...
 
Administrator
Joined 2004
Paid Member
I haven't yet heard a real, non-magical, argument in favour of passive crossovers.
I don't know what you've heard, but you should have read it in this thread several times.

See posts 100 and 108. They sum it up well.

I have another "non-magical" reason. I don't like the hassle of active. I like to swap amps and listen to them full range. I can't listen to an amp project full range on an active system - unless I have 3 identical.

My "magical" reason? Passive sounds just as good, if not better than active. Different, yes. But not worse.

Want more magic? Think harmonics.
 
Ex-Moderator R.I.P.
Joined 2005
So let's stick them in a computer and come up with the truly optimum passive crossover and marvel at its beauty.

But passive electronics is all about conflicting requirements and compromises (an 8th order passive is practically impossible, we're told, for example), so the digital active has simply got to be better in strictly objective terms.

I would stick to active :rolleyes:
 
This is nonsense.

At high level you can choose your best amplifier, best preamp, best speaker, best digital source. Those who work in audio shop selling high end stuffs will be familiar with the sound of certain high end system combinations. And one of these will be the ultimate passive system.

Now lets see the ultimate active analog crossover system... Let's list them.

As simple as that...
 
The crossover filters are strictly linear phase and complementary so if you add their outputs together the result is the original signal.

This in and of itself is useless for crossover design, it's nice in theory, but this only works if the drivers frequency responses are perfectly flat and extremely extended. 99.9% of drivers are not, so you will not be getting a flat response out of the loudspeakers, unless you've EQd them flat to start with.
 
This in and of itself is useless for crossover design, it's nice in theory, but this only works if the drivers frequency responses are perfectly flat and extremely extended. 99.9% of drivers are not, so you will not be getting a flat response out of the loudspeakers, unless you've EQd them flat to start with.

Hence my mention of further processing when I get round to it, and my suspicion that I've been a bit lucky with my choice of speakers (combined with room).
 
At high level you can choose your best amplifier, best preamp, best speaker, best digital source. Those who work in audio shop selling high end stuffs will be familiar with the sound of certain high end system combinations. And one of these will be the ultimate passive system.

Now lets see the ultimate active analog crossover system... Let's list them.

As simple as that...

This is just audiophile rubbish again. There are plenty of excellent active designs from a number of manufactures out there ATC being one of them. Just because the audiophile world shuns active loudspeakers in general doesn't mean they are automatically bad. The reason ironically why they shun them is because it tends to remove the part of picking your own parts. You tend to get a pair of loudspeakers with the amplifiers built in with no scope of 'upgrading' any of the parts in the future, it's boring for audiophiles for lack of a better word.

Interestingly though I have read the subjective reviews of many active systems over the years and they always get glowing reviews. Funny that.
 
OK, so you are not looking for objective reasons but more personal reasons. I'll tell you why I went back to using passives.

Actives are a big hassle when you are testing and using a high-end amp. You would have to use chip amps (Class D or gainclone) to keep costs and space usage down. My favorite amp is a single ended direct heated triode, and it is magical. But not practical to have three of these thing$ for active. Pretty much all modern recordings are close-mic'ed and are not supposed to have any sense of space or soundstage. It's very unnatural and "unmagical". You don't want hi-fi, you want realism, and the monotonically decreasing harmonics produced by a carefully tuned tube amp adds back the perception of distance and space. This is exactly the same thing air does to a sound wave, it adds harmonics. So now when I listen to my recordings, I close my eyes and feel like I am actually there. Now if I was using active, it would be my 6-channel T-amp, which is not the same.
 
Last edited:
In reply to Jay:


The big ATCs are pretty good.
I've heard them in a few studios and they come as close to the actual instrument as anything.
I've also heard very good things about ME Geithains RL901k from people I trust in that regard but they have eluded me as yet.

I'm quite happy with my own consisting of Fountek CD3.0 ribbons, Tannoy 3148 for treble and mids and Volt RV3143 in a T/L for bass. Amps are from the top: Omniphonics Footprint, MC2 Audio T500, MC2 Audio MC450 and MC2 Audio MC750.
Crossovers are modified BSS FDS360, I removed the limiters and relays and changed the op amps to OPA2134.
I use an ARX MultiQ parametric to replace eq function of the passive Tannoy crossover on the treble.
 
Just another Moderator
Joined 2003
Paid Member
WinISD Pro has a neat feature where you can add series resistance and watch the effect on box alignment - its worth (when using passive XOs, or long speaker cables), adding the resistance of the added components in, so the cabinet alignment can be altered accordingly.

Yes and in a lot of cases the response will be better.... Look at some of the older vifa woofers datasheets and you will see that they show their modelled responses with series resistance added (in the case of my M26WR-09-08's they show 0.7 ohms!) If the driver has been designed assuming that it will be used with a passive crossover so that that estimated series resistance will be present then you have to wonder, will you get the optimum performance if it is not present?

Tony.
 

Attachments

  • m26wr09-08.png
    m26wr09-08.png
    49.8 KB · Views: 143
Excuse my ignorance, but how does the designer of a passive crossover account for the changing electrical parameters (resistance, inductance?) of a drive unit depending on how hard it's being driven? Won't such changes affect crossover performance?

Yes it will and there's nothing they can do. The best thing possible is for the designer to see what the response looks like when the coils have warmed up a little and alter the component values perhaps to ensure that under a decent drive level things remain linear. For example if when driven hard a little a small hump appears around the xover frequency that could make the loudspeakers more forward and fatiguing then they should be, they could alter the xover values to compensate, but at the expense of flatness when the loudspeaker is cold. This could potentially make the loudspeaker sound good at lower levels, but also sound nice at higher levels too.

Typically though most people listen at levels that tend to stay on the cool side of things.
 
Excuse my ignorance, but how does the designer of a passive crossover account for the changing electrical parameters (resistance, inductance?) of a drive unit depending on how hard it's being driven? Won't such changes affect crossover performance?

Yes, they can, and do. But in practice, if the drivers are fairly good quality, and the crossover is well chosen in relation to the driver limitations, the effect of these changes can be fairly minimal.

That being said, it's easier to get truly high performance out of active systems than passive. In particular, what can be achieved with a DEQX is pretty magical if you start with the right speaker topologies and drivers.

But then, a nice full-range, passive XO with a bass augmentation driver can be pretty great also...

Yes, it's possible to achieve very high quality with both. Just more likely with the active.
 
Last edited:
This is just audiophile rubbish again. There are plenty of excellent active designs from a number of manufactures out there ATC being one of them.

Why rubbish? What I asked is to list the systems. You perceive audiophiles as probably some DIY-ers you know who do not know the knowledge nor the money.

Many "audiophiles" do not even care with DIY. They will buy anything that they like, whether it is active or passive. If you mention the best active system that will outperform their passive system, they will purchase it right away, no doubt.

But many audiophiles love the magic of simple tube amps. You wont see OPA2134 in their systems. Tell me which active analog crossovers that will not change the basic characteristics of 300B for example.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.