Active vs passive crossover?

[shellshocked]

Please forgive me but I am wrestling with active versus passive crossover design. If you put aside the need for additional amplifier channels are there advantages to passive versus active crossovers? There seems to be almost no mention of active crossovers on this forum. My assumption is that there must be a significant sound quality advantage to a well designed passive crossover but I don't know what it would be - am I missing something?

[KeithM]

Form my knowledge, most active xovers have to be designed to work with any speaker, so that means no EQ on peaks, no impedance leveling, etc.

[BobEllis]

Keith - what you say used to be true of commercial units, but there is significant EQ available in commercial designs such as the Behringer DCX2496. You can easily add whatever compensation your drivers need if you build your own. See http://www.linkwitzlab.com/ for lots of good info on active crossovers. SL even offers boards to allow you to build your own.

Passive XOs and the combined filter/driver responses can be modeled in freeware like speaker workshop. I haven't found an equivalent for active filters. I understand some of the more expensive commercial products can model active XOs. That's one strike against active.

If you already have enough amplifiers, active XOs can be less expensive to implement and tweak. I prefer active XOs for this reason and the improved control over the drivers resulting from direct connection. Most high end commercial speakers are passively crossed over primarily to allow the customer to purchase one amp per channel. Obviously, good results can be achieved either way, but I prefer active. Others will disagree, but I am not going back for speakers I plan to keep.

[peterS]

my vote is towards active

here is a x/o with much more credibility than any behringer product

http://www.123dj.com/crossovers/rane/ac23.html

http://www.rane.com/ac23b.html

rane is very good and well respected in profesional circles
behringer is not... for some reason it has a following on some internet forums
while i have not used it i trust the opinion of those around me to spare myself the trouble....
great features but bad sound from what i have gathered from multiple people in pro sound/ speaker manufacturing/ studio industries

again if you can spair the extra money go with rane, as i plan to do
123dj is a good source that i have used in the past for pro audio gear
call them up, they may quote a better price :T

[peterS]

and for those who's preamps dont have xlr outputs there are converters for composit to balanced.... no idea on price as thats a few perchases away yet

[ThomasW]

One doesn't need a fancy crossover for a sub. 24dB L/R designs work very well.

The primary benefit to active filters with a sub is not having the huge inductors in the signal path.

I recently obtained one of those Behringer crossovers that seem to lack credibility (CX2310). I bought it to put in the system temporarily while I repaired and upgraded my Marchand units. The Behringer is the sonic equal of anything I've ever used. And having had actively XO'ed subs since the late 1970's, I've had my fair share of experience with many different brands and custom made all discrete units I built and one JonMarsh built.

[peterS]

One doesn't need a fancy crossover for a sub. 24dB L/R designs work very well.

The primary benefit to active filters with a sub is not having the huge inductors in the signal path.

I recently obtained one of those Behringer crossovers that seem to lack credibility (CX2310). I bought it to put in the system temporarily while I repaired and upgraded my Marchand units. The Behringer is the sonic equal of anything I've ever used. And having had actively XO'ed subs since the late 1970's, I've had my fair share of experience with many different brands and custom made all discrete units I built and one JonMarsh built.

you're not going to hear any noise in a subwoofer anyways due to its inductance... its perfectly plausable it is a very good fit for that purpose
i am just hesitant to use one as i would intend... but like i said i have never used them

[Dennis H]

here is a x/o with much more credibility than any behringer product

http://www.123dj.com/crossovers/rane/ac23.html

Um, errr, that Rane model is fine for what it does, which is almost nothing toward developing a good-sounding fullrange speaker system. It's just a basic LR4 crossover with no baffle step, no customizing for driver response, no EQ to suppress cone resonances, no etc, etc. It's fine for a sub crossover, pretty worthless otherwise. When Rane sells an analog processor that can do what Siegfried Linkwitz (the king of active analog crossovers) does with his custom analog crossovers, I'll be impressed.



On the other hand, the Rane RPM digital models are fine units that can do anything you want. The CAD-style signal path design is way cool.

"Credibility" is just a bunch of people flapping their gums and patting themselves on the back.

[peterS]

also worth noting that given a large enough amplifier, biwiring passives will yeild equvilent sq improvements

#84 on www.bcae1.com explains to some extent

[ThomasW]

also worth noting that given a large enough amplifier, biwiring passives will yeild equvilent sq improvements

This is false.....

you're not going to hear any noise in a subwoofer anyways due to its inductance... its perfectly plausable it is a very good fit for that purpose

That would be correct except for the fact that the system I'm using it in is tri-amped.

[peterS]

This is false.....

i see...
care to help me out in my other thread

[peterS]

Um, errr, that Rane model is fine for what it does, which is almost nothing toward developing a good-sounding fullrange speaker system. It's just a basic LR4 crossover with no baffle step, no customizing for driver response, no EQ to suppress cone resonances, no etc, etc. It's fine for a sub crossover, pretty worthless otherwise. When Rane sells an analog processor that can do what Siegfried Linkwitz (the king of active analog crossovers) does with his custom analog crossovers, I'll be impressed.



On the other hand, the Rane RPM digital models are fine units that can do anything you want. The CAD-style signal path design is way cool.

"Credibility" is just a bunch of people flapping their gums and patting themselves on the back.

So you think the behrengers are good?
Any idea on the cost on the rane rpm, and more importantly what they do 8O

[Dennis H]

also worth noting that given a large enough amplifier, biwiring passives will yeild equvilent sq improvements

Oh boy, where to begin? Somebody just doesn't get it.

[Dennis H]

so you think the behrengers are good?
any idea on the cost on the rane rpm, and more importantly what they do 8O

Google is your friend.

[peterS]

Oh boy, where to begin? Somebody just doesn't get it.

guess i dont get the attitude... guess you were just born knowing everthing.... you guys realy know how to suck the fun out of this

[Dennis H]

guess i dont get the attitude

Sorry if I offended. I detected a bit of "attitude" along with some vigorously asserted but flat wrong information in your postings to this thread. Just trying to nip the misinformation in the bud.

[peterS]

no attitude from me till i started receiving it... im here to learn

so what what do you guys suggest for a triamped setup

time alignment, eq , and ofcourse x/o

and what is the consensus on behrenger's eq's....

[ThomasW]

Peter,

no attitude from me till i started receiving it... im here to learn

Adult world here no blame...You'll find that Bob Ellis (who posted earlier) has been designing and building loudspeakers since the 1970's. JonMarsh and I have been designing and building loudspeaker as a team since 1969. And Dennis, well let's just say he knows a lot about all things audio ...😉

So you think the behrengers are good?

We don't 'think' they're good, we know they are for home audio use. And we know they aren't worth squat for use in the touring pro audio world

Any idea on the cost on the rane rpm, and more importantly what they do

Yes they start about $600 and go up to approx $2700 depending on the amount of processing you want. And yes they're programmed like most DSP audio devices. If you compare specs you'll find the Behringer DCX2496 has similar features for a fraction the price. And they're used by some people with pretty high end systems like Victor's massive dipole:

[thylantyr]

Another vote for DCX2496. I use it, I've tried all digital path,
analog path, etc. I can't make this unit sound bad :T ..
and for $250 shipped, what a bargain. If you are new
or a seasoned pro, you can benefit that that digital crossover,
it's a great tool to have in your bag O' tricks. nothing up my
sleeve.. presto! :lol:

[shellshocked]

Thanks to everyone for the excellent response. I was leaning active and this has pushed me over the edge.

[ThomasW]

Thanks to everyone for the excellent response. I was leaning active and this has pushed me over the edge.

Actually your thread was hijacked and as a result you've probably drawn the wrong conclusion.

A well designed/engineered passive crossover is frequently the best choice particularly if you're building something like a relatively simple 2 or 3-way.

Active XO's mean things like baffle step compensation and other things needed to customize the speaker's operations must be done with active circuitry, and frequently that's more easily done with passive circuitry.

So the choice of active vs passive is related to the specific design being created.

[JonMarsh]

Take a close look at the block diagram for the Orion crossover- it has all of the kinds of compensation that might be built into a passive crossover, plus some more for the dipole behavior. This can be done active in a digital, but the thing to watch out for with the less expensive digitals is that you want to keep them operating with lots of bits- which in practice means feeding them with a moderately high level signal, then using a multiple section attenuator afterwards as the volume control.

Also, the digital's tend not to have the required dynamic range for dipole bass EQ if you take it to the lengths SL does in his active analog crossover- if you're doing a conventional enclosure type speaker, then you won't have that worrry.

Often the most practical approach is a combination of passive and active, using the latter for the lower frequency crososvers where less driver compensation may be needed.

~Jon

[AJINFLA]

As an avowed follower of the active approach for years now, I should add that you can also do both analog and digital active in a hybrid setup. Or passive/active bass if you want to keep things simpler. I'll readily admit that Jons passive designs cede nothing to any active ones. It's only in the bass region where the active really has a clear advantage (such things as servo control of the bass driver - that I also favor ).
But keep in mind that even Jons terrific designs are geared more for a static setup. Someone who wants to build something permanent, one shot deal, with maybe a little tweaking of components here or there. One of the advantages (for the hopeless tweaker like I) of DSP is not only endless adjustability, but something almost impossible to do passively - active room adapatability after building, through eq,etc so moving your speaker in room or even to another room becomes non too detrimental.
As noted before, it comes down to (just like speaker building) the end user properly implementing the equipment at their disposal.
I own both a Rane AC22 and a DCX2496. Both can be transparent in your music chain, if used properly. Remember, the W22 can be made to sound bad. Just don't blame the driver :W .
My next design probably will use the DCX for M/T xo and eq, with the Rane for bass.
Keep that option in mind also. Along with a passive elliptic/active bass type approach.

Cheers,

AJ

[Dennis H]

And Dennis, well let's just say he knows a lot about all things audio ...:wink:

Thanks for the vote of confidence, Thomas.

At times, I feel like a bit of poser here because I'm not currently building anything. The last few years, my music/HT system has been a 27" TV and headphones because of family circumstances. I moved in with my 90-year-old parents to look after them and keep them out of a nursing home. But I do have a strong DIY background. In the 60's a highschool pal and I started building speakers, amps, etc. much like you and Jon. He was the electronics guy and I was the cabinet guy with a smattering of electronics. He made a career of the electronics bit, opening a stereo repair business that later morphed into a high-end retail store. I started out in medical electronics, got bored and confused, and started building houses for a living. But through it all we never lost our interest in DIY audio.

Anyway, thanks to everybody here at HTGuide for helping to keep me sane while I'm on a self-imposed DIY hiatus. I'm living vicariously through you guys. ;x(

[JonMarsh]

Thanks for the vote of confidence, Thomas.

At times, I feel like a bit of poser here because I'm not currently building anything. The last few years, my music/HT system has been a 27" TV and headphones because of family circumstances. I moved in with my 90-year-old parents to look after them and keep them out of a nursing home. But I do have a strong DIY background. In the 60's a highschool pal and I started building speakers, amps, etc. much like you and Jon. He was the electronics guy and I was the cabinet guy with a smattering of electronics. He made a career of the electronics bit, opening a stereo repair business that later morphed into a high-end retail store. I started out in medical electronics, got bored and confused, and started building houses for a living. But through it all we never lost our interest in DIY audio.

Anyway, thanks to everybody here at HTGuide for helping to keep me sane while I'm on a self-imposed DIY hiatus. I'm living vicariously through you guys. ;x(

Well, hat's off to you for what you're doing, Dennis. I know of friends of friends that have been in similar situations. We'll try to crank up the action this year so there will be more fun things to read about and look forward to.

Be sure to drop by around the end of March now and then.

[JonMarsh]

As an avowed follower of the active approach for years now, I should add that you can also do both analog and digital active in a hybrid setup. Or passive/active bass if you want to keep things simpler. I'll readily admit that Jons passive designs cede nothing to any active ones. It's only in the bass region where the active really has a clear advantage (such things as servo control of the bass driver - that I also favor ).
But keep in mind that even Jons terrific designs are geared more for a static setup. Someone who wants to build something permanent, one shot deal, with maybe a little tweaking of components here or there. One of the advantages (for the hopeless tweaker like I) of DSP is not only endless adjustability, but something almost impossible to do passively - active room adapatability after building, through eq,etc so moving your speaker in room or even to another room becomes non too detrimental.
As noted before, it comes down to (just like speaker building) the end user properly implementing the equipment at their disposal.
I own both a Rane AC22 and a DCX2496. Both can be transparent in your music chain, if used properly. Remember, the W22 can be made to sound bad. Just don't blame the driver :W .
My next design probably will use the DCX for M/T xo and eq, with the Rane for bass.
Keep that option in mind also. Along with a passive elliptic/active bass type approach.

Cheers,

AJ

You're right on the money about my INTENTIONS; to have finished stable systems. Alas, it's not always what I manage, especially not the more "aggressive" projects. I've been contemplating getting a DCX2496 just as a developement tool for that reason, just as soon as something comes up I'm convinced it's a must have for. Right now, I'm working on getting setup for a measurement binge. Later this year, it will probably be a development and building binge. Some pretty silly things I'm thinking about... :B but they should be fun, and that's why I do this.

[noah katz]

Very interesting discussion.

I'm a bit surprised that passive still holds so much sway.

I had been under the impression that one of the biggest advantages of active over passive was the ability to do very steep slopes, like 48 dB/oct, with the attendant benefits of lower allowable XO, and less driver overlap and lobing.

Perhaps I need to revise this view, especially in light of what I've been reading about Jon's XO's.

Have any of you heard the NHT Xd?

Thanks

[cotdt]

Has anyone ever made crossovers from light bulbs and gyrators?

[JonMarsh]

Very interesting discussion.

I'm a bit surprised that passive still holds so much sway.

I had been under the impression that one of the biggest advantages of active over passive was the ability to do very steep slopes, like 48 dB/oct, with the attendant benefits of lower allowable XO, and less driver overlap and lobing.

Perhaps I need to revise this view, especially in light of what I've been reading about Jon's XO's.

Have any of you heard the NHT Xd?

Thanks

Yup, but I'm not a major fan of the DEQX or switching amplifiers in the full audio spectrum. Just an old Luddite at heart. I researched digital techniques starting over 25 years ago, and the last paper I published on a switching amp design was an AES paper in 1991. The DEQX is currently the last word in off the shelf Digital crossover and equalization systems, with a price tag to match; but it's not the last word in sonic transparency. In many cases the total system design can be improved considerably with the DEQX, but at some point I think it will be a bottle neck. YMMV, check out the various reviews available and listen to one yourself before drawing any firm conclusions or spending your hard earned cash.

While you can implement a true LR-8 crossover easily with one, for example, that has more group delay and to my ears doesn't sound as good as the CE 8th order approach I use, which has about the same GD as an LR4.

If you want to understand more about my approach to electronics, read up about Ayre products and designs. Their VP of engineering used to work for me in the late 70's while he was still a physics student at CU. Chas and I are pretty aligned in our experiences and opinions about what works for ultimate quality, and what doesn't. But hey, it's free country, and I recommend everyone experiment and determine what they like.

What's that old saying about if you meet the Buddha on the road, kill him?

That's really just a way of saying you have to travel the path yourself to get to where YOU need to be, and do the work yourself, within reason.

~Jon

[AJINFLA]

Has anyone ever made crossovers from light bulbs and gyrators?

No, but I have made a shoe stink.
Seriously though, I think the old light bulb/gyrator crossovers are strictly for use with waveguide loaded plasma tweeters. Cotdt, if you ever get around to building a contraption such as this, please do post some pics.
That would truly be a sight to behold.

Cheers,

AJ

[Dennis H]

While you can implement a true LR-8 crossover easily with one, for example, that has more group delay and to my ears doesn't sound as good as the CE 8th order approach I use, which has about the same GD as an LR4.

Oh boy, taking a deep breath and risking Jon sending Evil Twin to pay me a visit......

The DEQX uses linear-phase FIR filters so there is no group delay other than a frequency-independent processing-time delay. DEQX crossovers of whatever slope can play a nice square wave. Of course the bandpass behavior of the drivers will degrade that somewhat but it's still pretty dang good in that regard. That's what sets the DEQX (and the TacT) above other digital XO boxes.

[Dennis H]

That said, occasional contributor MAC built the Orions, sold 'em and tried to "improve" on 'em with the DCX, DEQX and various drivers, gave up, and built his second set of Orions for his listening enjoyment.

[JonMarsh]

Evil Twin is on vacation this weekend, but my bad, you're right about the FIR filters.

OTOH., I've heard comments from several people that I fairlly respect that the overall character doesn't pass a straight wire bypass kind of test.

Then, of course there is that minor detail that a DEQX costs more than I've spent for drivers, passive crossover parts, and an active crossover on my most expensive system to day, the X1 SLAMM Klones. Apart from that, it's an interesting concept, and if I could buy one for $500 to $1K, I'd probably be more receptive. More thoughts later- I mean, if you guys want to take up a collection and get me one for Christmas this year, I'd be sure to give it a real good wring out and develop a few systems with it! :B

Besides, there has to be one Luddite hold-out just to make things interesting around here, doesn't there?

[noah katz]

Jon,

"While you can implement a true LR-8 crossover easily with one, for example, that has more group delay and to my ears doesn't sound as good as the CE 8th order approach I use, which has about the same GD as an LR4."

Your passive XO's *8th* order?! I don't remember seeing that many components; or do you mean 8th order acoustic, 6th order electrical?

"I've heard comments from several people that I fairlly respect that the overall character doesn't pass a straight wire bypass kind of test."

Now I'm confused; have you heard a digital XO yourself?

Thanks

[JoshK]

I myself own the DEQX and I wouldn't say that what Jon or Dennis said was off-base. It is a very expensive toy, but I felt it offered enough to make it worthwhile for me. I guess though that I am not a luddite.

The DEQX isn't perfect and I am sure it isn't the last word in transparency, however I know AudioEngr has remarked that it could be incredibly transparent and world class with some mods (expensive ones!!). I being a modder might try it out myself at some time down the road when that becomes the bottleneck in my system. Currently it isn't.

SL has some good comments about the benefits and tradeoffs of DSP xo's, including iirc FIR ones. There is the pre-ringing artifact. Ultimately I know a number of pro audio speakers that use DSP xo's suggest minimum phase settings instead of linear phase settings. I think I'll ultimate try ot mimic Jon's transfer function via the DEQX and if I'm successful, I'll compare to other settings.

[Dennis H]

Chuckle, I was looking back at this old thread and remembering something Jon said: we'd know he was having a midlife crisis if he started trading stock index futures. Well, I guess I'm in full crisis mode. The relevance is I had to turn my meager analog electronics knowledge into digital code when I started trading -- a good job to do from home if you can actually make money at it. And after I studied a bunch and figured all that out on trading data, I realized the same stuff is being done in audio these days. Same old DSP code, same old streaming data. Of course, Josh is the pro at this stuff, I'm just an amateur scalping a little along the edges of what the big boys are doing.

[JoshK]

which Josh are you referring to?

[Dennis H]

You.... aren't you the quant guy?

[Victor]

Well, I guess I am going to jump right in, - the original question in this thread was ”… are there advantages to passive versus active crossovers…” This is an important question and it needs to be addressed in the fundamental manner.

Passive crossover consists of the energy storing elements, such as potentially big caps and inductors. The power amplifier must charge them, i.e. store energy. This energy will then have to be used. However, it takes time for this energy to do work, - hence the problem. The amplifier is trying to control the speaker, but the energy storing elements in the crossover prevent the amplifier from accomplishing this control in a timely manner. The analogy might be this, - the amplifier-passive crossover-speaker system acts much like a truck that you are trying to stop quickly or accelerate quickly, but you can’t due to the inherent inertia. In electrical terms this produces unwanted distortion.

Active crossover has none of the problems associated with the energy storage because the amplifier speaker connection is direct and, in this case, the amplifier has a near complete control of the speaker cone, so when the amplifer ‘says’ stop the cone stops right away, or when the amp ‘says’ go, - the cones responds immediately. This way we have much better control and a very little distortion.

Naturally, there is no ‘free launch’ in engineering and the price to pay for the benefits of the active cross-over are extra amplifiers, although with lower power settings and possibly better THD.

A lot has been said about the use of digital cross-overs, particularly the DCX496. I use 2 of them with my 4 channel dipole set-up. Thomas in his post #18 kindly referenced my front channels. They are indeed very large. Due to the nature of the dipole EQ I ended up using an analog circuit to EQ the woofers. The DCX does not have the required headroom. I have absolutely no concerns about the signal transparency of the DCX, - it works very well indeed and it replaces a good 2 dozens op-amps that the analog active crossover would have had. Let along the unbelievable controllability that it offers is along worth it for me. I use a fully digital signal chain including the volume controls, so the input dynamic range is of no concern to me, but if the analog inputs are used with the DSP-type crossover then a great care must be taken to make sure that the input amplitude is close to 0db level.

There was a mention of the fact that the rather expensive DEQX crossover uses FIR filters which are desirable for audio due to their linear phase characteristics. The DCX2496 uses the IIR filters which have the same phase characteristics as their analog counterparts. It is erroneous to say that FIR filters would necessarily reproduce better square wave. It depends on the number of coefficients used in the FIR filter. Gibbs Phenomena and the associated passband ringing is a part of the FIR filter reality, - good square wave is only possible if the FIR filter uses a very large number of coefficients. On the other hand IIR filters may have a very nice phase characteristics and good looking square waves if Linkwitz-Ralley filter alignment is used. But then again, - who cares for sharp square waves in audio anyway, - a slow slope is just fine.

Well, having said all that I am now working on a DIY DSP crossover solution for my dipoles that will feature FIR filters.

Regards,
Victor

[Dennis H]

"Energy storage" is simply a matter of the transfer function, i.e. the frequency response of the filter. It's identical whether you do it with small caps in a high-impedance active circuit or with big caps in a low-impedance passive circuit.

Linkwitz-Riley lowpass plus highpass circuits give an allpass sum which cannot accurately reproduce a square wave.

[JoshK]

You.... aren't you the quant guy?

Yes, I am. There is another Josh on this forum, and I wasn't sure if you were referring to me. But yes, I am a quant in the finance industry. I guess I didn't remember saying that. Good memory.

[Victor]

"Energy storage" is simply a matter of the transfer function, i.e. the frequency response of the filter. It's identical whether you do it with small caps in a high-impedance active circuit or with big caps in a low-impedance passive circuit.

That might be, - but with a passive crossover the impedance of the speaker itself is a factor, - so the energy that the big caps store become problematic. Furthermore, the linear behavior of the big caps is an approximation at best. There is just no way to achieve the same cone control with the passive crossover that you can do with an active circuit.

Linkwitz-Riley lowpass plus highpass circuits give an allpass sum which cannot accurately reproduce a square wave.

Well, - nothing can reproduce the square wave accurately. The L-R approach is good enough in this regard. We are only concerned with below 20 kHz here, so a slow slewing square wave is fine, as long as the settling is well behaved and your graph shows that.

[opt-e]

I am planning on building a pair of sealed Nat P's and matching sealed passive RSS315 bass bins which will be able to act as stands. I've been looking for an active crossover solution to high pass the MTM's and lowpass the bass bins. I'm a stereo only guy and I don't have a reciever that can provide this functionality.

Would something like the Behringer CX2310 be a good fit, or should I spring for the DCX2496? I am also open to other suggestions.

[ThomasW]

The CX2310 is a very good unit. I picked one up a few months back as a temp fix while I repaired my Marchand's. It's still in the system because there was no audible difference between it and the Marchand (and other projects like this weeks installation of a new toilet have gotten in the way........ :roll: )

Just know that the cx2310 in/outputs are XLR only. They can be run single ended by making up some custom cables

[opt-e]

Thomas, are you using the unit as a low-pass for your sub only, or as a high-pass to your mains as well?

The other consideration is that I want to do some equalization on my bass bins to boost the low end a bit and compensate for any room modes. Would the DCX2496 be able to do this or would it be better to get the CX as a crossover and a separate BFD for bass equalization?

I'm straying a bit off topic now, perhaps I should start a new thread.